HIV-1 subtyping using phylogenetic analysis of pol gene sequences

Journal of Virological Methods 94 (2001) 45 – 54 www.elsevier.com/locate/jviromet

HIV-1 subtyping using phylogenetic analysis of pol gene sequences C. Pasquier *, N. Millot, R. Njouom, K. Sandres, M. Cazabat, J. Puel, J. Izopet Laboratoire de Virologie, Hoˆpital Purpan, Place Baylac, Toulouse, France Received 19 October 2000; received in revised form 22 January 2001; accepted 23 January 2001

Abstract HIV-1 pol gene sequencing is now used routinely in France to identify mutations associated with resistance to reverse transcriptase (RT) or protease (PR) inhibitors. These sequences may also provide other information, such as the HIV-1 subtype. HIV-1 subtyping was compared using the RT and PR gene sequences to heteroduplex mobility assay (HMA) of the envelope gene. The RT and PR genes of 51 samples that had been subtyped earlier by HMA were sequenced. Sequences were aligned and subtypes were determined by phylogenetic analysis with reference HIV sequences. HMA gave the following subtypes: A (20), B (19), C (1), D (3), F (1), G (3) and CRF01-AE (4). Phylogenetic analysis of the RT gene gave: A (5), B (19), C (2), D (3), F (1), G (6), J (2), CRF01 – AE (4), CFR02 – AG (7) and undetermined (2). PR gene analysis did not infer subtypes with sufficient confidence. HMA and RT subtyping was not in agreement in nine cases. RT subtyping can identify CFR02 – AG and CRF01 – AE variants from A subtype RT. It was shown that phylogenetic analysis of the RT gene could provide a useful method for HIV-1 subtyping. The length of the amplicon and the relative performance of each primer pair used in this study favoured RT sequences as a subtyping tool. One potential advantage over en6 subtyping HMA is the ability to identify some circulating recombinant forms (CRFs). © 2001 Elsevier Science B.V. All rights reserved. Keywords: HIV-1; Gene; Phylogenetic analysis

1. Introduction HIV-1 is known for its remarkable genetic variability. HIV-1 variants are randomly generated during virus replication and then selected by the host environment. Two mechanisms are responsible for producing virus variants. One is the errorprone nature of the reverse transcriptase (RT), * Corresponding author.

which has no proofreading function and causes nucleotide substitutions, deletions and insertions (Preston and Dougherty, 1996; Mansky, 1998). The second is the recombination generated by the low processivity of RT and the presence of two copies of genomic RNA in the nucleocapsid (Quinones-Mateu and Arts, 1999). Both mechanisms intervene during HIV-1 replication, which occurs at high rate (1010 viral particles produced each day) and continues for long periods within

0166-0934/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 6 - 0 9 3 4 ( 0 1 ) 0 0 2 7 2 - 5

C. Pasquier et al. / Journal of Virological Methods 94 (2001) 45–54

46

individuals and human populations (Ho et al., 1995; Perelson et al., 1997). The newly produced variants survive only if they can compete with other variants to replicate and escape immune responses and antiretroviral molecules. The fittest variant for a specific micro environment is then selected and may become a major variant. HIV-1 variants has been divided into groups (M, N and O) and subtypes; A, B, C, D, F1, F2, G, H, J, K for group M based on phylogenetic analyses of complete genome nucleotide sequences (Robertson et al., 2000). Intersubtype recombinant viruses have been identified and recently named as CRF01 to CRF06 (circulating recombinant forms (CRFs)) (Robertson et al., 2000). HIV-1 group M viruses are distributed world-wide and are the most common in developed countries. The distribution of group subtypes is both geographical and epidemiological. HIV-1 subtypes are relevant epidemiological tools for studying the evolution of the epidemic. In those studies, HIV-1 subtypes are often determined using V3 serotyping or en6 heteroduplex mobility assays (Arens, 1999). The en6 sequences are used preferentially for subtyping because of their great variability. Nucleotide sequences from other HIV-1 genes that also vary significantly can also be useful for subtyping HIV-1 and can be of value for identifying recombinant genomes. The development of RT and protease (PR) gene sequencing for searching for mutations conferring resistance to RT and PR-inhibitors has made nucleotide sequences from those regions available as an alternative method for subtyping HIV-1. Subtyping of HIV-1 was compared using phylogenetic analysis of RT and PR nucleotide sequences with that using a heteroduplex mobility assay on en6 sequences, which is a method used widely. Samples giving discordant results were also analysed by sequencing the en6 region.

2. Materials and methods

2.1. Samples A total of 51 HIV-1 strains from patients liv-

ing in the Toulouse area were used. These strains had been subtyped using the heteroduplex mobility assay (HMA) to identify non-B subtype HIV1 infections in several settings. First, they were used for patients (n= 21) with differences in plasma HIV-1 RNA concentration by bDNA and RT-PCR Monitor assays, or by Monitor v1.0 and v1.5 assays (Roche Diagnostic, Meylan France). Second, they were used for patients (n = 18) with a low CD4 cell count and low and stable plasma HIV-1 RNA load. Last, they were used for patients from sub-Saharan countries (n = 12). Plasma was prepared by centrifugation at 600×g for 10 min, and clarified by centrifugation for 15 min at 3000× g to insure cell-free specimens; it was stored at − 80°C. Citrated peripheral blood samples were centrifuged over Lymphocyte Separation Medium (Organon Teknika, USA) density gradients. Peripheral blood mononuclear cells (PBMCs) were washed twice with phosphate-buffered saline (PBS), and counted. Five million PBMCs were pelleted, dried and stored at − 80°C.

2.2. Heteroduplex mobility assay HIV-1 subtypes were identified using the heteroduplex mobility assay (HMA) (Delwart et al., 1993). Briefly, each PBMC pellet was lysed for 2 h at 56°C in 10 mM Tris–HCl (pH 8.5), 50 mM KCl, 2.5 mM MgCl2, 0.45% NP40, 0.45% Tween 20 and 80 m/ml proteinase K. The proteinase K was then inactivated by heating the mixture for 2 min at 96°C. A 600 base-pair fragment (spanning the V3–V5 en6 gene region) was amplified by nested PCR from 2 mg of HIV-1-infected PBMC DNA. The same 600 bp fragment was amplified from each of the reference plasmids containing en6 gene sequences from the various HIV-1 subtypes (A–H) (Delwart et al., 1995). The PCR product from each strain was mixed, in separate tubes, with an equal quantity of the PCR product of each of the reference plasmid sequences and denaturated. The mixture was loaded onto a non-denaturing acrylamide gel to study the migration of the molecular hybrid obtained (heteroduplex).

C. Pasquier et al. / Journal of Virological Methods 94 (2001) 45–54

2.3. En6 gene sequencing

47

Courtaboeuf, France) and sequenced on both strands by the dideoxy chain termination method. All the measures to prevent contamination suggested by Kwok and Higushi (1989) were applied.

The PCR product used for HMA was sequenced using ES7 and ES8 primers. The PCR products were purified on QIAmp columns (Qiagen, Courtaboeuf, France) and sequenced on both strands by the dideoxy chain termination method (ABI PRISM Ready Reaction AmpliTaq FS, Dye DeoxyTerminator, Applied Biosystems, Paris, France) on an ABI377 automated DNA sequencer (Applied Biosystems).

The sequences were submitted to EMBL with accession no. AF330740–AF 330790 for RT sequences, AF330707–AF330739 for PR sequences and AF335755–AF335764 for en6 sequences.

2.4. RT and PR gene sequencing

2.6. Analysis of sequence data

The reverse transcriptase and protease genes were sequenced as described earlier (Izopet et al., 1998). Briefly, plasma HIV-1 RNA was extracted using the Qiamp Viral RNA kit (QIAGEN, Courtaboeuf, France). HIV-1 RNA was reverse transcribed at 37°C for 60 min using 20 U of M-MuLV reverse transciptase (Boehringer GmbH, Mannheim, Germany) and 10 mM of antisense outer primers (RT2 or PR2). RT amplification used the outer primers RT2 and RT1 and the inner primers RT3 and RT4. PR amplification was carried out with PR2 and PR1 as outer primers and PR3 and PR4 as inner primers. PCR was carried out in 10 mM Tris– HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 0.2 mM of each deoxynucleoside triphosphate, 50 pM of each primer, 2.5 U Taq polymerase (AmpliTaq, Perkin Elmer Cetus, Norwalt, CT) and 5 ml cDNA solution. The final volumes were 50 ml for the first round of PCR and 100 ml for the second round of PCR. The primary PCR involved initial denaturation at 94°C for 5 min; 35 cycles of denaturation at 94°C for 60 s; annealing at 55°C for 60 s, and polymerisation at 72°C for 150 s with a final elongation at 72°C for 10 min. An aliquot (5 ml) of the primary PCR products was used for 35 cycles of nested PCR as follows; initial denaturation at 94°C for 5 min, 35 cycles of denaturation at 94°C for 30 s, annealing at 55°C for 30 s, and polymerisation at 72°C for 1 min, with a final elongation at 72°C for 5 min. The RT3-RT4 amplification gave a 774-bp product and the PR3PR4 amplification a 366-bp product. These were purified on QIAmp columns (Qiagen,

All sequences were checked for possible contamination using the protocol recommended by Kuiken and Korber (http://hiv-web.lanl.gov/ HTML/Contam/contam – main.html). Multiple alignments were done with Sequence Navigator (Perkin Elmer Applied Biosystems, USA) and CLUSTALW version 1.7 (32) programs. The alignment was adjusted by hand before phylogenetic analysis with version 3.572c of the Phylogeny Inference Package (PHYLIP). Phylogenetic distances between sequences were calculated using the two-parameter Kimura method (DNADIST from PHYLIP) with a transition-transversion ratio of 2.0. Dendograms were created by the Neighbor-Joining and Maximum Likelihood methods with CLUSTALW and PHYLIP programs. Tree diagrams were plotted with the TREEVIEW version 1.6 program. Bootstrapping was performed on the Neighbor-Joining tree using CLUSTALW 1.7.

2.5. Nucleotide sequence accession numbers

2.7. Statistical analysis The  2 test or the Fisher’s exact test was used to compare distribution ratios. The Wilcoxon test was used to compare bootstrap values.

3. Results

3.1. HMA subtyping HIV-1 isolates were subtyped by heteroduplex mobility assay using a 600 bp DNA fragment

48

C. Pasquier et al. / Journal of Virological Methods 94 (2001) 45–54

overlapping the V3 region. As indicated in Table 1, HIV-1 subtypes were B (n =19), A (n = 20), C (n = 1), D (n=3), F (n = 1), G (n = 3) and CRF01-AE (n= 4).

3.2. RT gene subtyping A 774 bp from RT gene was amplified, sequenced directly and subtyped by phylogenetic analysis using reference sequences and the Neighbor-Joining method (Fig. 1). The RT gene was amplified in all cases. The HIV-1 subtype was determined in 47/51 (92%) strains. Subtypes were B (n = 19), A (n=5), C (n =2), D (n = 3), F (n= 1), G (n= 6), CRF01 – AE (n =4), CFR02 – AG (n= 7) and J (n =2). Two strains (sequences c14 and c 16) had undetermined subtypes using Neighbor-Joining, maximum likelihood or maximum parsimony phylogenetic analysis. Analysis of these sequences showed no sign of recombination between subtypes. If the two related J strains are not considered, since there is no corresponding reference plasmid available for HMA, then HMA and RT subtyping agreed for 41/49 subtypes (84%). We considered strains that were A using HMA and CRF02-AG using RT subtyping to be in agreement, since CRF02 – AG strains are reported to be subtyped A using en6 gene (Kuiken et al., 1999). The two isolates with unidentified subtypes by RT sequence analysis were subtyped A by HMA. The 11 sequences with subtype mismatches were A (n = 10) and B (n=1) using HMA. All discordant cases were confirmed by analysis of en6 and RT gene sequences obtained by direct sequencing of PCR products. The results obtained using en6 sequencing were concordant with the HMA subtyping (Fig. 3). This points to recombination between subtypes for the en6 and RT genes.

3.3. PR gene subtyping A 366 bp from the PR gene was amplified, sequenced directly and subtyped as described for the RT gene. The PR gene was amplified in only 33 out of 51 cases (64%), this is not significantly lower than for the RT gene (P = 0.17). Amplifica-

tion of PR was less successful in non-B subtypes than in B subtype (PB 0.001). The HIV-1 subtype was determined in 31 out of 33 strains. Bootstrap values were lower for the PR tree than for the RT tree (PB 0.05) (Fig. 2). The subtypes were B (n= 19), A (n= 2), D (n = 1), G (n = 2), CRF01 – AE (n= 3) and CFR02 – AG (n = 4). Two strains (sequences c 8 and c41) had undetermined subtypes using Neighbor-Joining, maximum likelihood and maximum parsimony phylogenetic analyses (Fig. 3). Subtyping based on en6 HMA and PR sequences was concordant for 26 subtypes (79%). Four isolates had discordant subtypes. Two isolates subtyped F and G by HMA were subtyped B using PR phylogenetic analysis. Two isolates subtyped A by HMA were G using PR sequences. The RT and PR sequences analyses were different for four subtypes. They were C/G, CRF02 – AG/B, F/B and G/B. Strain c 03 was found to be related to three different subtypes using the three different methods.

4. Discussion The use of HIV-1 RT and PR genotyping for determining resistance mutations will produce a large amount of pol gene sequence data. These available sequences may provide complementary information to physicians, particularly for HIV-1 subtypes. This study assesses the performance of HIV-1 subtyping using RT and PR sequences for phylogenetic analysis, using HMA, a technique, which is used widely, as reference. The set of strains used in this study is not representative of the distribution of HIV-1 subtypes in France. Non-B subtypes are much rarer, at 15–20% according to a national survey done in 1996–1998 (Couturier et al., 2000). The RT sequence gave consistent HIV-1 subtypes in most cases, and identified some recombinant strains. CRF02 – AG formed a different subtree from the G sequences, as well as CRF06. CRF01 – AE strains, known to be E in the en6 region and A in the pol and gag genes, clustered in a different subtree than the A subtype reference strains. Nevertheless, there were differences be-

C. Pasquier et al. / Journal of Virological Methods 94 (2001) 45–54

49

Fig. 1. Phylogenetic analysis of RT sequences of HIV-1. A Neighbor-Joining phylogenetic tree was built from isolates and reference RT sequences from the Los Alamos HIV data base (http://hiv-web.lanl.gov). The Kimura two-parameter method of estimating genetic distance was used. Numbers next to the nodes of the tree represent bootstrap values (1000 replicates).

50

C. Pasquier et al. / Journal of Virological Methods 94 (2001) 45–54

Fig. 2. Phylogenetic analysis of PR sequences of HIV-1. A Neighbor-Joining phylogenetic tree was built from isolates and reference PR sequences from the Los Alamos HIV data base (http://hiv-web.lanl.gov). The Kimura two-parameter method of estimating genetic distance was used. Numbers next to the nodes of the tree represent bootstrap values (1000 replicates).

C. Pasquier et al. / Journal of Virological Methods 94 (2001) 45–54

51

Fig. 3. Phylogenetic analysis of C2V3 en6 sequences of HIV-1. A Neighbor-Joining phylogenetic tree was built from isolates and reference sequences from the Los Alamos HIV data base (http://hiv-web.lanl.gov). The Kimura two-parameter method of estimating genetic distance was used. Numbers next to the nodes of the tree represent bootstrap values (1000 replicates) (Arens, 1999).

C. Pasquier et al. / Journal of Virological Methods 94 (2001) 45–54

52

Table 1 HIV-1 subtyping using en6 HMA and phylogenetic analysis of RT and PR sequences Patient

En6 region HMA

RT sequence

PR sequence

Patient origin

01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

A A A A A A A A A A A A A A A A A A A A B B B B B B B B B B B B B B B B B B B C D D D CRF01-AE CRF01-AE CRF01-AE CRF01-AE F G G G

G CRF02-AG C G A A A J J CRF02-AG CRF02-AG G CRF02-AG Ua A Ua A B CRF02-AG CRF02-AG B B CRF02-AG B B B B B B B B B B B B B B B B C D D D CRF01-AE CRF01-AE CRF01-AE A/CRF01-AE F G G G

– CRF02-AGa Ga Ga – CRF01-AEa Aa Ua – – – – CRF02-AGa – Aa – – – CRF02-AGa CRF02-AGa Ba Ba Ba Ba Ba Ba Ba Ba Ba – Ba – Ba Ba Ba Ba Ba Ba Ba – Ua – Da CRF01-AEa CRF01-AEa – – Ba Ba – –

Cameroon – Ivory Coast Democratic Republic of Congo Democratic Republic of Congo France Democratic Republic of Congo France France Ivory Coast France France – Gabon France France France – France Burkina Faso France France Cameroon France Spain France France France France Spain France France France France France France France France France France Central African Republic France France France – Thailand Central African Republic Gabon France France France

a

U indicates unknown subtype, number of bootstraps below 600 for 1000 replicates.

C. Pasquier et al. / Journal of Virological Methods 94 (2001) 45–54

tween the three methods used on a single isolate. Sequencing and phylogenetic analysis of en6 PCR products were used for HMA to eliminate possible misinterpretation of HMA in these cases, and it was concordant with the HMA results. The lack of subtype J and CRFs in the HMA reference plasmid panel could also give false HMA results. This was not the case in our study, since en6 sequences confirmed the HMA results. These disagreements could be due to inter-subtype recombination between envelope, RT and PR genes. Inter-subtype recombinants were probably underestimated, because of the almost exclusive use of the C2V3 en6 region for subtyping. The use of full genome subtyping has led to the identification of several recombinant strains with sometimes complex mosaic genome structures (Salminen et al., 1995; Gao et al., 1998). An alternative to full HIV-1 genome sequencing is to study multiple genome regions in a single isolate. This approach usually uses HMA (Heyndrickx et al., 2000) or sequencing (Cornelissen et al., 1996; Liitsola et al., 1998; Morris et al., 1999). Therefore, the use of available RT and PR sequences can easily complement en6 genotyping without requiring new manipulations. Since the RT gene varies less than the en6 gene, sequence analysis of some subtypes or CRFs cannot discriminate fully between strains using RT sequences. For example, CRF05 and subtype F, F1 and F2 subtypes are not separated clearly by phylogenetic analysis of the RT gene. This may also be due to the fewer sequences available for the RT region compared with the en6 region. As the two RT sequences whose subtype was not determined had no mutation associated with resistance and since some subtyped RT sequences had up to six mutations, the presence of resistance mutations does not seem to significantly influence subtyping. PR phylogenetic analysis appears to be less powerful than RT sequence analysis. PCR amplification and sequencing were less successful and the bootstrap values were below the confidence threshold. This is probably due to the size of the PR sequence, which are only one half of the RT gene fragment analysed, and possibly fewer variations between subtypes. Mutations associated with resistance to protease inhibitors may also be in-

53

volved by increasing sequence variability without respect for subtype specificity. Since the same plasma sample was used for both RT and PR PCR amplifications, the level of HIV-1 viral load was not involved in the absence of PR amplification. New primers should be developed to increase the frequency of PR gene amplification, in particular for non-B subtypes. In summary, we have shown that phylogenetic analysis of RT gene could offer a useful method for HIV-1 subtyping. The length of the amplicon and the relative performance of each primer pair used in this study favoured RT sequences as a subtyping tool. One potential advantage over en6 subtyping HMA is its ability to identify some CRFs.

Acknowledgements We thank Dr Owen Parkes and Monica Ghosh for linguistic advice.

References Arens, M., 1999. Methods for subtyping and molecular comparison of human viral genomes. Clin. Microbiol. Rev. 12 (4), 612 – 626. Cornelissen, M., Kampinga, G., Zorgdrager, F., Goudsmit, J., 1996. Human immunodeficiency virus type 1 subtypes defined by en6 show high frequency of recombinant gag genes. J. Virol. 70 (11), 8209 – 8212. Couturier, E., Damond, F., Roques, P., Fleury, H., Barin, F., J-B, B., Brun-Ve´ zinet, F., Simon, F., the AC11 laboratory network, 2000. HIV-1 diversity in France, 1996-1998. AIDS 14(3), 289 – 296. Delwart, E.L., Herring, B., Rodrigo, A.G., Mullins, J.I., 1995. Genetic subtyping of human immunodeficiency virus using a heteroduplex mobility assay. Pcr. Methods Appl. 4 (5), S202 – S216. Delwart, E.L., Shpaer, E.G., Louwagie, J., McCutchan, F.E., Grez, M., Ruebsamen, W.H., Mullins, J.I., 1993. Genetic relationships determined by a DNA heteroduplex mobility assay: analysis of HIV-1 env genes. Science 262 (5137), 1257 – 1261. Gao, F., Robertson, D.L., Carruthers, C.D., Morrison, S.G., Jian, B., Chen, Y., Barre, S.F., Girard, M., Srinivasan, A., Abimiku, A.G., Shaw, G.M., Sharp, P.M., Hahn, B.H., 1998. A comprehensive panel of near-full-length clones and reference sequences for non-subtype B isolates of human immunodeficiency virus type 1. J. Virol. 72 (7), 5680 – 5698.

54

C. Pasquier et al. / Journal of Virological Methods 94 (2001) 45–54

Study Group on Heterogeneity of HIV Epidemics in African Cities, Heyndrickx, L., Janssens, W., Zekeng, L., Musonda, R., Anagonou, S., Van der Auwera, G., Coppens, S., Vereecken, K., De Witte, K., Van Rampelbergh, R., Kahindo, M., Morison, L., McCutchan, F.E., Carr, J.K., Albert, J., Essex, M., Goudsmit, J., Asjo, B., Salminen, M., Buve, A., van Der Groen, G., 2000. Simplified strategy for detection of recombinant human immunodeficiency virus type 1 group M isolates by gag/env heteroduplex mobility assay. J. Virol. 74 (1), 363 – 370. Ho, D.D., Neumann, A.U., Perelson, A.S., Chen, W., Leonard, J.M., Markowitz, M., 1995. Rapid turnover of plasma virions and CD4 lymphocytes in HIV-1 infection. Nature 373 (6510), 123 –126. Izopet, J., Salama, G., Pasquier, C., Sandres, K., Marchou, B., Massip, P., Puel, J., 1998. Decay of HIV-1 DNA in patients receiving suppressive antiretroviral therapy. J. Acquir. Immun. Defic. Syndr. Hum. Retrovirol. 19, 478 –483. Kuiken, C., Foley, B., Hahn, B., Marx, P., McCutchan, F., Mellors, J., Mullins, J., Wolinsky, S., Korber, B., 1999. Human Retrovirus and AIDS -A compilation and analysis of nucleic acid and amino acid sequences [http://hivweb.lanl.gov]. Los Alamos National Laboratory. Kwok, S., Higushi, R., 1989. Avoiding false positives with PCR. Nature 339, 227 –238. Liitsola, K., Tashkinova, I., Laukkanen, T., Korovina, G., Smolskaja, T., Momot, O., Mashkilleyson, N., Chaplinskas, S., Brummer, K.H., Vanhatalo, J., Leinikki, P., Salminen, M.O., 1998. HIV-1 genetic subtype A/B recom-

.

binant strain causing an explosive epidemic in injecting drug users in Kaliningrad. AIDS 12 (14), 1907 – 1919. Mansky, L., 1998. Retrovirus mutation rates and their role in genetic variation. J. Gen. Virol. 79, 1337 – 1345. Morris, A., Mardsen, M., Halcrow, K., Hughes, E., Brettle, R., Bell, J., Simmonds, P., 1999. Mosaic structure of the human immunodeficiency virus type 1 genome infecting lymphoid cells and the brain: evidence for frequent in vivo recombination events in the evolution of regional populations. J. Virol. 73 (10), 8720 – 8731. Perelson, A.S., Essunger, P., Ho, D.D., 1997. Dynamics of HIV-1 and CD4+ lymphocytes in vivo, AIDS Suppl. S17 – 24. Preston, B., Dougherty, J., 1996. Mechanisms of retroviral mutation. Trends Microbiol. 4 (1), 16 – 22. Quinones-Mateu, M., Arts, E., 1999. Recombination in HIV1: update and implications. AIDS Rev. 1, 89 – 100. Robertson, D.L., Anderson, J.P., Bradac, J.A., Carr, J.K., Foley, B., Funkhouser, R.K., Gao, F., Hahn, B.H., Kalish, M.L., Kuiken, C., Learn, G.H., Leitner, T., McCutchan, F., Osmanov, S., Peeters, M., Pieniazek, D., Salminen, M., Sharp, P.M., Wolinsky, S., Korber, B., 2000. HIV-1 nomenclature proposal. Science 288 (5463), 55 – 56. Salminen, M.O., Koch, C., Sanders, B.E., Ehrenberg, P.K., Michael, N.L., Carr, J.K., Burke, D.S., McCutchan, F.E., 1995. Recovery of virtually full-length HIV-1 provirus of diverse subtypes from primary virus cultures using the polymerase chain reaction. Virology 213 (1), 80 – 86.