hiv-1 subtyping using phylogenetic analysis of pol … subtyping using phylogenetic analysis of pol...

Journal of Virological Methods 94 (2001) 45–54

HIV-1 subtyping using phylogenetic analysis of pol genesequences

C. Pasquier *, N. Millot, R. Njouom, K. Sandres, M. Cazabat, J. Puel,J. Izopet

Laboratoire de Virologie, Hopital 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 toreverse transcriptase (RT) or protease (PR) inhibitors. These sequences may also provide other information, such asthe HIV-1 subtype. HIV-1 subtyping was compared using the RT and PR gene sequences to heteroduplex mobilityassay (HMA) of the envelope gene. The RT and PR genes of 51 samples that had been subtyped earlier by HMAwere sequenced. Sequences were aligned and subtypes were determined by phylogenetic analysis with reference HIVsequences. 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 andRT subtyping was not in agreement in nine cases. RT subtyping can identify CFR02–AG and CRF01–AE variantsfrom A subtype RT. It was shown that phylogenetic analysis of the RT gene could provide a useful method for HIV-1subtyping. The length of the amplicon and the relative performance of each primer pair used in this study favouredRT sequences as a subtyping tool. One potential advantage over en� subtyping HMA is the ability to identify somecirculating recombinant forms (CRFs). © 2001 Elsevier Science B.V. All rights reserved.

Keywords: HIV-1; Gene; Phylogenetic analysis

www.elsevier.com/locate/jviromet

1. Introduction

HIV-1 is known for its remarkable genetic vari-ability. HIV-1 variants are randomly generatedduring virus replication and then selected by thehost environment. Two mechanisms are responsi-ble for producing virus variants. One is the error-prone nature of the reverse transcriptase (RT),

which has no proofreading function and causesnucleotide substitutions, deletions and insertions(Preston and Dougherty, 1996; Mansky, 1998).The second is the recombination generated by thelow processivity of RT and the presence of twocopies of genomic RNA in the nucleocapsid(Quinones-Mateu and Arts, 1999). Both mecha-nisms intervene during HIV-1 replication, whichoccurs at high rate (1010 viral particles producedeach day) and continues for long periods within* Corresponding author.

0166-0934/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.

PII: S 0166 -0934 (01 )00272 -5

C. Pasquier et al. / Journal of Virological Methods 94 (2001) 45–5446

individuals and human populations (Ho et al.,1995; Perelson et al., 1997). The newly producedvariants survive only if they can compete withother variants to replicate and escape immuneresponses and antiretroviral molecules. The fittestvariant for a specific micro environment is thenselected 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 phylogeneticanalyses of complete genome nucleotide se-quences (Robertson et al., 2000). Intersubtyperecombinant viruses have been identified and re-cently named as CRF01 to CRF06 (circulatingrecombinant forms (CRFs)) (Robertson et al.,2000). HIV-1 group M viruses are distributedworld-wide and are the most common in devel-oped countries. The distribution of group sub-types is both geographical and epidemiological.HIV-1 subtypes are relevant epidemiologicaltools for studying the evolution of the epidemic.In those studies, HIV-1 subtypes are often deter-mined using V3 serotyping or en� heteroduplexmobility assays (Arens, 1999).

The en� sequences are used preferentially forsubtyping because of their great variability. Nu-cleotide sequences from other HIV-1 genes thatalso vary significantly can also be useful for sub-typing HIV-1 and can be of value for identifyingrecombinant genomes. The development of RTand protease (PR) gene sequencing for searchingfor mutations conferring resistance to RT andPR-inhibitors has made nucleotide sequencesfrom those regions available as an alternativemethod for subtyping HIV-1.

Subtyping of HIV-1 was compared using phy-logenetic analysis of RT and PR nucleotide se-quences with that using a heteroduplex mobilityassay on en� sequences, which is a method usedwidely. Samples giving discordant results werealso analysed by sequencing the en� 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 strainshad been subtyped using the heteroduplex mobil-ity assay (HMA) to identify non-B subtype HIV-1 infections in several settings. First, they wereused for patients (n=21) with differences inplasma HIV-1 RNA concentration by bDNAand RT-PCR Monitor assays, or by Monitorv1.0 and v1.5 assays (Roche Diagnostic, MeylanFrance). Second, they were used for patients(n=18) with a low CD4 cell count and low andstable plasma HIV-1 RNA load. Last, they wereused for patients from sub-Saharan countries(n=12). Plasma was prepared by centrifugationat 600×g for 10 min, and clarified by centrifu-gation for 15 min at 3000×g to insure cell-freespecimens; it was stored at −80°C. Citrated pe-ripheral blood samples were centrifuged overLymphocyte Separation Medium (OrganonTeknika, USA) density gradients. Peripheralblood mononuclear cells (PBMCs) were washedtwice with phosphate-buffered saline (PBS), andcounted. Five million PBMCs were pelleted,dried and stored at −80°C.

2.2. Heteroduplex mobility assay

HIV-1 subtypes were identified using the het-eroduplex mobility assay (HMA) (Delwart et al.,1993). Briefly, each PBMC pellet was lysed for 2h at 56°C in 10 mM Tris–HCl (pH 8.5), 50 mMKCl, 2.5 mM MgCl2, 0.45% NP40, 0.45% Tween20 and 80 �/ml proteinase K. The proteinase Kwas then inactivated by heating the mixture for 2min at 96°C. A 600 base-pair fragment (spanningthe V3–V5 en� gene region) was amplified bynested PCR from 2 �g of HIV-1-infected PBMCDNA. The same 600 bp fragment was amplifiedfrom each of the reference plasmids containingen� gene sequences from the various HIV-1 sub-types (A–H) (Delwart et al., 1995). The PCRproduct from each strain was mixed, in separatetubes, with an equal quantity of the PCRproduct of each of the reference plasmid se-quences and denaturated. The mixture wasloaded onto a non-denaturing acrylamide gel tostudy the migration of the molecular hybrid ob-tained (heteroduplex).

C. Pasquier et al. / Journal of Virological Methods 94 (2001) 45–54 47

2.3. En� gene sequencing

The PCR product used for HMA was se-quenced using ES7 and ES8 primers. The PCRproducts were purified on QIAmp columns (Qia-gen, Courtaboeuf, France) and sequenced on bothstrands by the dideoxy chain termination method(ABI PRISM Ready Reaction AmpliTaq FS, DyeDeoxyTerminator, Applied Biosystems, Paris,France) on an ABI377 automated DNA se-quencer (Applied Biosystems).

2.4. RT and PR gene sequencing

The reverse transcriptase and protease geneswere sequenced as described earlier (Izopet et al.,1998). Briefly, plasma HIV-1 RNA was extractedusing the Qiamp Viral RNA kit (QIAGEN,Courtaboeuf, France). HIV-1 RNA was reversetranscribed at 37°C for 60 min using 20 U ofM-MuLV reverse transciptase (BoehringerGmbH, Mannheim, Germany) and 10 �M ofantisense outer primers (RT2 or PR2). RT am-plification used the outer primers RT2 and RT1and the inner primers RT3 and RT4. PR amplifi-cation was carried out with PR2 and PR1 as outerprimers and PR3 and PR4 as inner primers. PCRwas carried out in 10 mM Tris–HCl (pH 8.3), 50mM KCl, 1.5 mM MgCl2, 0.2 mM of each de-oxynucleoside triphosphate, 50 pM of eachprimer, 2.5 U Taq polymerase (AmpliTaq, PerkinElmer Cetus, Norwalt, CT) and 5 �l cDNA solu-tion. The final volumes were 50 �l for the firstround of PCR and 100 �l for the second round ofPCR. The primary PCR involved initial denatura-tion at 94°C for 5 min; 35 cycles of denaturationat 94°C for 60 s; annealing at 55°C for 60 s, andpolymerisation at 72°C for 150 s with a finalelongation at 72°C for 10 min. An aliquot (5 �l)of the primary PCR products was used for 35cycles of nested PCR as follows; initial denatura-tion at 94°C for 5 min, 35 cycles of denaturationat 94°C for 30 s, annealing at 55°C for 30 s, andpolymerisation at 72°C for 1 min, with a finalelongation at 72°C for 5 min. The RT3-RT4amplification gave a 774-bp product and the PR3-PR4 amplification a 366-bp product. These werepurified on QIAmp columns (Qiagen,

Courtaboeuf, France) and sequenced on bothstrands by the dideoxy chain termination method.All the measures to prevent contamination sug-gested by Kwok and Higushi (1989) were applied.

2.5. Nucleotide sequence accession numbers

The sequences were submitted to EMBL withaccession no. AF330740–AF 330790 for RT se-quences, AF330707–AF330739 for PR sequencesand AF335755–AF335764 for en� sequences.

2.6. Analysis of sequence data

All sequences were checked for possible con-tamination using the protocol recommended byKuiken and Korber (http://hiv-web.lanl.gov/HTML/Contam/contam–main.html). Multiplealignments were done with Sequence Navigator(Perkin Elmer Applied Biosystems, USA) andCLUSTALW version 1.7 (32) programs. Thealignment was adjusted by hand before phyloge-netic analysis with version 3.572c of the Phy-logeny Inference Package (PHYLIP).Phylogenetic distances between sequences werecalculated using the two-parameter Kimuramethod (DNADIST from PHYLIP) with a transi-tion-transversion ratio of 2.0. Dendograms werecreated by the Neighbor-Joining and MaximumLikelihood methods with CLUSTALW andPHYLIP programs. Tree diagrams were plottedwith the TREEVIEW version 1.6 program. Boot-strapping was performed on the Neighbor-Joiningtree using CLUSTALW 1.7.

2.7. Statistical analysis

The �2 test or the Fisher’s exact test was usedto compare distribution ratios. The Wilcoxon testwas used to compare bootstrap values.

3. Results

3.1. HMA subtyping

HIV-1 isolates were subtyped by heteroduplexmobility assay using a 600 bp DNA fragment


overlapping the V3 region. As indicated in Table1, HIV-1 subtypes were B (n=19), A (n=20), C(n=1), D (n=3), F (n=1), G (n=3) andCRF01-AE (n=4).

3.2. RT gene subtyping

A 774 bp from RT gene was amplified, se-quenced directly and subtyped by phylogeneticanalysis using reference sequences and the Neigh-bor-Joining method (Fig. 1). The RT gene wasamplified in all cases. The HIV-1 subtype wasdetermined in 47/51 (92%) strains. Subtypes wereB (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 c16) hadundetermined subtypes using Neighbor-Joining,maximum likelihood or maximum parsimonyphylogenetic analysis. Analysis of these sequencesshowed no sign of recombination betweensubtypes.

If the two related J strains are not considered,since there is no corresponding reference plasmidavailable for HMA, then HMA and RT subtypingagreed for 41/49 subtypes (84%). We consideredstrains that were A using HMA and CRF02-AGusing RT subtyping to be in agreement, sinceCRF02–AG strains are reported to be subtypedA using en� gene (Kuiken et al., 1999). The twoisolates with unidentified subtypes by RT se-quence analysis were subtyped A by HMA. The11 sequences with subtype mismatches were A(n=10) and B (n=1) using HMA. All discordantcases were confirmed by analysis of en� and RTgene sequences obtained by direct sequencing ofPCR products. The results obtained using en�sequencing were concordant with the HMA sub-typing (Fig. 3). This points to recombination be-tween subtypes for the en� and RT genes.

3.3. PR gene subtyping

A 366 bp from the PR gene was amplified,sequenced directly and subtyped as described forthe RT gene. The PR gene was amplified in only33 out of 51 cases (64%), this is not significantlylower than for the RT gene (P=0.17). Amplifica-

tion of PR was less successful in non-B subtypesthan in B subtype (P�0.001). The HIV-1 subtypewas determined in 31 out of 33 strains. Bootstrapvalues were lower for the PR tree than for the RTtree (P�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 c8 and c41) had undetermined sub-types using Neighbor-Joining, maximum likeli-hood and maximum parsimony phylogeneticanalyses (Fig. 3).

Subtyping based on en� HMA and PR se-quences was concordant for 26 subtypes (79%).Four isolates had discordant subtypes. Two iso-lates subtyped F and G by HMA were subtyped Busing PR phylogenetic analysis. Two isolates sub-typed A by HMA were G using PR sequences.

The RT and PR sequences analyses were differ-ent for four subtypes. They were C/G, CRF02–AG/B, F/B and G/B. Strain c03 was found to berelated to three different subtypes using the threedifferent methods.

4. Discussion

The use of HIV-1 RT and PR genotyping fordetermining resistance mutations will produce alarge amount of pol gene sequence data. Theseavailable sequences may provide complementaryinformation to physicians, particularly for HIV-1subtypes. This study assesses the performance ofHIV-1 subtyping using RT and PR sequences forphylogenetic analysis, using HMA, a technique,which is used widely, as reference.

The set of strains used in this study is notrepresentative of the distribution of HIV-1 sub-types in France. Non-B subtypes are much rarer,at 15–20% according to a national survey done in1996–1998 (Couturier et al., 2000).

The RT sequence gave consistent HIV-1 sub-types in most cases, and identified some recombi-nant strains. CRF02–AG formed a differentsubtree from the G sequences, as well as CRF06.CRF01–AE strains, known to be E in the en�region and A in the pol and gag genes, clusteredin a different subtree than the A subtype referencestrains. Nevertheless, there were differences be-


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


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


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


Table 1HIV-1 subtyping using en� HMA and phylogenetic analysis of RT and PR sequences

PR sequence Patient originRT sequencePatient En� region HMA

01 G – CameroonACRF02-AGa –CRF02-AGA02

CA Ga Ivory Coast03Ga04 Democratic Republic of CongoA G– Democratic Republic of CongoA05 ACRF01-AEa France06 A AAa Democratic Republic of CongoA07 A

JA Ua France08– FranceJA09

CRF02-AGA – Ivory Coast10–11 FranceA CRF02-AG– FranceGA12CRF02-AGa –13 A CRF02-AG– GabonUa14 A

AA Aa France15– FranceUaA16

AA – France17– –18 A BCRF02-AGa FranceCRF02-AG19 ACRF02-AGa Burkina Faso20 A CRF02-AGBa FranceB21 B

BB Ba France22Ba CameroonCRF02-AGB23

BB Ba France24Ba25 SpainB BBa FranceB26 BBa France27 B BBa FranceB28 B

BB Ba France29– SpainBB30

BB Ba France31– France32 B BBa FranceB33 BBa France34 B BBa FranceB35 B

BB Ba France36Ba FranceBB37

BB Ba France38Ba France39 B B– FranceCC40Ua Central African Republic41 D D– FranceD42 D

DD Da France43CRF01-AEa FranceCRF01-AECRF01-AE44

CRF01-AECRF01-AE CRF01-AEa –45– Thailand46 CRF01-AE CRF01-AE– Central African RepublicA/CRF01-AE47 CRF01-AEBa Gabon48 F FBa FranceG49 G

GG – France50– FranceG51 G

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


tween the three methods used on a single isolate.Sequencing and phylogenetic analysis of en� PCRproducts were used for HMA to eliminate possiblemisinterpretation of HMA in these cases, and it wasconcordant with the HMA results. The lack ofsubtype J and CRFs in the HMA reference plasmidpanel could also give false HMA results. This wasnot the case in our study, since en� sequencesconfirmed the HMA results.

These disagreements could be due to inter-sub-type recombination between envelope, RT and PRgenes. Inter-subtype recombinants were probablyunderestimated, because of the almost exclusive useof the C2V3 en� region for subtyping. The use offull genome subtyping has led to the identificationof several recombinant strains with sometimescomplex mosaic genome structures (Salminen et al.,1995; Gao et al., 1998). An alternative to full HIV-1genome sequencing is to study multiple genomeregions in a single isolate. This approach usuallyuses HMA (Heyndrickx et al., 2000) or sequencing(Cornelissen et al., 1996; Liitsola et al., 1998;Morris et al., 1999). Therefore, the use of availableRT and PR sequences can easily complement en�genotyping without requiring new manipulations.

Since the RT gene varies less than the en� gene,sequence analysis of some subtypes or CRFs can-not discriminate fully between strains using RTsequences. For example, CRF05 and subtype F, F1and F2 subtypes are not separated clearly byphylogenetic analysis of the RT gene. This may alsobe due to the fewer sequences available for the RTregion compared with the en� region.

As the two RT sequences whose subtype was notdetermined had no mutation associated with resis-tance and since some subtyped RT sequences hadup to six mutations, the presence of resistancemutations does not seem to significantly influencesubtyping.

PR phylogenetic analysis appears to be lesspowerful than RT sequence analysis. PCR amplifi-cation and sequencing were less successful and thebootstrap values were below the confidencethreshold. This is probably due to the size of thePR sequence, which are only one half of the RTgene fragment analysed, and possibly fewer varia-tions between subtypes. Mutations associated withresistance to protease inhibitors may also be in-

volved by increasing sequence variability withoutrespect for subtype specificity. Since the sameplasma sample was used for both RT and PR PCRamplifications, the level of HIV-1 viral load was notinvolved in the absence of PR amplification. Newprimers should be developed to increase the fre-quency of PR gene amplification, in particular fornon-B subtypes.

In summary, we have shown that phylogeneticanalysis of RT gene could offer a useful method forHIV-1 subtyping. The length of the amplicon andthe relative performance of each primer pair usedin this study favoured RT sequences as a subtypingtool. One potential advantage over en� subtypingHMA is its ability to identify some CRFs.

Acknowledgements

We thank Dr Owen Parkes and Monica Ghoshfor linguistic advice.

References

Arens, M., 1999. Methods for subtyping and molecular com-parison 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 subtypesdefined by en� show high frequency of recombinant gaggenes. J. Virol. 70 (11), 8209–8212.

Couturier, E., Damond, F., Roques, P., Fleury, H., Barin, F.,J-B, B., Brun-Vezinet, F., Simon, F., the AC11 laboratorynetwork, 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 usinga 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. Geneticrelationships determined by a DNA heteroduplex mobilityassay: 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 andreference sequences for non-subtype B isolates of humanimmunodeficiency virus type 1. J. Virol. 72 (7), 5680–5698.


Study Group on Heterogeneity of HIV Epidemics in AfricanCities, Heyndrickx, L., Janssens, W., Zekeng, L., Mu-sonda, 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 strategyfor detection of recombinant human immunodeficiencyvirus type 1 group M isolates by gag/env heteroduplexmobility assay. J. Virol. 74 (1), 363–370.

Ho, D.D., Neumann, A.U., Perelson, A.S., Chen, W., Leon-ard, J.M., Markowitz, M., 1995. Rapid turnover of plasmavirions and CD4 lymphocytes in HIV-1 infection. Nature373 (6510), 123–126.

Izopet, J., Salama, G., Pasquier, C., Sandres, K., Marchou, B.,Massip, P., Puel, J., 1998. Decay of HIV-1 DNA inpatients receiving suppressive antiretroviral therapy. J. Ac-quir. 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 analysisof nucleic acid and amino acid sequences [http://hiv-web.lanl.gov]. Los Alamos National Laboratory.

Kwok, S., Higushi, R., 1989. Avoiding false positives withPCR. Nature 339, 227–238.

Liitsola, K., Tashkinova, I., Laukkanen, T., Korovina, G.,Smolskaja, T., Momot, O., Mashkilleyson, N., Chaplin-skas, 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 injectingdrug users in Kaliningrad. AIDS 12 (14), 1907–1919.

Mansky, L., 1998. Retrovirus mutation rates and their role ingenetic 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 thehuman immunodeficiency virus type 1 genome infectinglymphoid cells and the brain: evidence for frequent in vivorecombination events in the evolution of regional popula-tions. J. Virol. 73 (10), 8720–8731.

Perelson, A.S., Essunger, P., Ho, D.D., 1997. Dynamics ofHIV-1 and CD4+ lymphocytes in vivo, AIDS Suppl.S17–24.

Preston, B., Dougherty, J., 1996. Mechanisms of retroviralmutation. Trends Microbiol. 4 (1), 16–22.

Quinones-Mateu, M., Arts, E., 1999. Recombination in HIV-1: 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-1nomenclature 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 ofdiverse subtypes from primary virus cultures using thepolymerase chain reaction. Virology 213 (1), 80–86.

.

hiv-1 subtyping using phylogenetic analysis of pol … subtyping using phylogenetic analysis of pol...

Documents