Journal of Clinical Virology 30 (2004) 224–228

Characterization of human immunodeficiency virus type-1 from HIV-1seropositive cases with undetectable viremia

Chunfu Yang∗, Ming Li, Faye Cowart, Donna Rudolph, Bharat Parekh,J. Steve McDougal, Renu B. Lal

Division of AIDS, STD, and TB Laboratory Research, National Center for HIV, STD and TB Prevention, Centers for Disease Control and Prevention,Mail Stop D-12, 1600 Clifton Road, Atlanta, GA 30333, USA

Received 19 September 2003; received in revised form 27 October 2003; accepted 4 November 2003

Abstract

Background:Human immunodeficiency virus type 1 (HIV-1) viral load has become a standard of care among HIV-1-infected patients;however, a small number of patients have undetectable viral load even though they have never been treated.Methods:By using RT-PCRand DNA-PCR, and followed by sequencing and phylogenetic analyses, a detailed molecular characterization was carried out from fiveHIV-1-seropositive patients who had undetectable viral load by commercially available ultrasensitive viral load assays.Results:Of the fourpatients whose plasmas were available, viral RNAs were detected in three of them by using an in-house RT-PCR in at least one of the threeregions (integrase, protease orenvgp41). The fourth patient had positive RT-PCR signals in these regions only when RNA isolated from thesupernatant of cocultivated patient PBLs with PHA-stimulated HIV-1 negative donor PBLs was used. Further analysis of DNA extractedfrom the PBMCs revealed that four of the five patients had detectable proviral sequences in at least two of the three regions. The fifth patienthad only positive PCR results in all three regions when DNA isolated from PHA-stimulated patient’s PBLs was used. Phylogenetic analysisof protease andenvgp41 regions revealed that three patients were infected with subtype B viruses while the remaining two patients wereinfected with subtype C and CRF02AG viruses. These subtypes coincided with geographic origin and known molecular epidemiology ofHIV-1 infection. Conclusion:These data provide evidence that both subtype B and non-B HIV-1 infection can result in undetectable viralload in HIV-1-infected patients and that efforts should continue to further characterize these viruses.© 2003 Elsevier B.V. All rights reserved.

Keywords:HIV-1; Diagnostics; HIV-1 RNA or DNA

1. Introduction

Quantification of HIV-1 viral RNA has become a stan-dard tool for clinical management of HIV disease. Whereasviral load assays were initially developed to accuratelyquantify subtype B, the worldwide spread of non-B sub-

Abbreviations: HIV-1, human immunodeficiency virus type 1;RT-PCR, reverse transcriptase-polymerase chair reaction; DNA-PCR, de-oxyribonucleic acid-polymerase chair reaction;env, HIV envolope gene;gag, HIV group antigen;int, HIV integrase gene;pol, HIV polymerasegene;prt, HIV protease gene; SIVcpz, simian immunodeficiency virus ofchimpanzees; PBMC, peripheral blood mononuclear cells; PBL, periph-eral blood lymphocytes; PHA, phytohemagglutinin; gp41, glycoprotein 41;CRF, circulating recombinant form; RNA, ribonucleic acid; DNADIST,deoxyribonucleic acid distance

∗ Corresponding author. Tel.:+1-404-639-4975;fax: +1-404-639-2660.

E-mail address:cyang1@cdc.gov (C. Yang).

types, the migration of people infected with non-B subtypesto the USA, and the introduction of treatment programsin regions with non-B subtypes have raised concerns re-lated to accurate quantification of these viruses (Hu et al.,1998; Lucas et al., 1999; Sullivan et al., 2000). Generally,HIV-1-seropositive patients have detectable plasma viralRNA, unless they are on highly active antiretroviral therapy(HAART). Studies have shown that in individuals with suc-cessful HAART, plasma viral RNA levels are suppressedto undetectable levels with current methodology (Lucaset al., 1999). Two other scenarios with lack of detectableviral RNA involve a subset of HIV-infected individualswho are long-term non-progressors (Migueles et al., 2000;Propato et al., 2001) and those infected with an HIV-variantthat may be highly divergent and may not amplify withthe assay primers (Emery et al., 2000; Parekh et al.,1999).

1386-6532/$ – see front matter © 2003 Elsevier B.V. All rights reserved.doi:10.1016/j.jcv.2003.11.007

C. Yang et al. / Journal of Clinical Virology 30 (2004) 224–228 225

We have recently developed qualitative molecular detec-tion assays using generic primer pairs in theenvandpol re-gions (Yang et al., 1999, 2000). We have used these highlysensitive and broadly reactive detection assays in the molec-ular confirmation of HIV-1 infections from patients whowere serologically positive for HIV-1 infections but had un-detectable viral load by commercially available ultrasensi-tive load assays (detection limit 50 copies/ml of plasma).

2. Materials and methods

2.1. Patients

The patient samples tested in this study were either fromthe CDC HIV diagnostics reference laboratory (HIV Im-munology and Diagnostics Branch, National Center for HIV,STD, and TB Prevention) or from primary care physiciansrequesting confirmation of HIV infection. A total of fivesamples from patients infected with HIV-1 from the USA(n = 4) and New Zealand (n = 1) were obtained in 1999 and2000. All are HIV-1 seropositive, with no detectable plasmaviral RNA using various viral load assays. The table sum-marizes available demographic and clinical characteristicsof these patients. In brief, patient 1 is an HIV-1-seropositivemale living in New York City. He tested HIV-1 positive in1989, 1997, and 1999, but his CD4 counts have been nor-mal and he has no AIDS defining symptoms. Patient 2 isan HIV-1-seropositive young male from Guinea with AIDS.He has been on intermittent antiretroviral regimens; how-ever, his viral load is below detection limit by UltraSensitiveAmplicor HIV-1 MonitorTM, version 1.5 (Roche DiagnosticSystems, Inc. Branchburg, NJ, USA) during treatment andwithout treatment. Patient 3 is a 49-year-old male havingan injecting drug use history; however, the exact date of in-fection is not known. Patient 4 is a 36-year-old Sri Lankanrefugee who tested positive for HIV-1 antibodies in 1996 and2000, consistently with undetectable viral load. In addition,repeated in-house PCR assays at the referring laboratory us-ing primers from HIV-1gag (SKK38/39 and SK145/431)andenv(SK68/69) and from HIV-2 were all negative. Patient5 is a 75-year-old female widow who tested HIV-1 positiveduring a pre-surgery checkup. Look-back analysis revealedthat she had a blood transfusion in 1984, presumably result-ing in HIV-1 infection. She maintains normal CD4 countsand has no detectable viral load.

2.2. HIV serologic tests and viral load measurements

HIV serology tests were repeated at the CDC HIV di-agnostics reference laboratory with Genetic SystemTM

HIV-1/HIV-2 peptide EIA kit (Bio Rad Laboratories, Red-mond, WA, USA) and Cambridge Biotech HIV-1 Westernblot kit (Calypte Biomedical Corp., Rockville, MD, USA).The HIV-1 viral load was determined with UltraSensitiveAmplicor HIV-1 MonitorTM, version 1.5.

2.3. Amplification of HIV-1 viral sequences by RT-PCRand DNA-PCR

Viral RNA was extracted from 200�l of plasma usingQIAamp viral RNA kit (Qiagen, Avenue Stanford, Valencia,CA, USA) from all patients except patient 4 from whom aplasma sample was not available. Genomic DNA was alsoextracted from patient’s PBMCs using Qiagen DNA bloodmini kit (Qiagen, Avenue Stanford, Valencia, CA, USA).RT-PCR or DNA-PCR was then performed using genericprimers gpM-Z and intM-Z previously shown to amplify allknown HIV-1 viruses and SIVcpz (Yang et al., 1999, 2000).HIV-1 group M-specific and HIV-2-specific primers in theprotease region (Janini et al., 1996; Masciotra et al., 2002)were also used. Primer sequences and protocols have beendescribed in detail elsewhere (Janini et al., 1996; Masciotraet al., 2002; Yang et al., 1999, 2000).

2.4. Sequencing and phylogenetic analysis

The purified nested PCR products from selected PCR am-plicons were used for automated sequencing with BigDye

Table 1Demographic and clinical characteristics and PCR results of the patients

Patients

1 2 3 4 5

Sex/age (year) M/NA M/NA M/49 M/36 F/75

HIV testEIA + + + + +WB + + + + +VLa <50 <50 <50 <50 <50

CD4 (cells/l) Normal 98 350 1010 NormalRisk factors None None IDUb None BTc

ARV-Rx None Yes None None None

HIV-1 viral detectionRT-PCR

Int + + + ND +d

gp41 B ND B ND +d

Prt NDe ND B ND +d

DNA-PCRInt + − + + +d

gp41 B +f B C Bd

Prt B AG B − Bd

HIV-2 viral detectionRT-PCR–prt − − − ND −DNA-PCR–prt − − − − −a Viral load.b Injecting drug user.c Blood transfusion in 1984.d Patient 5 was positive forint, prt andenvgp41 regions when DNA iso-

lated from PHA-stimulated PBLs was used for DNA-PCR or RNA isolatedfrom the supernatant of cocultivated patient PBLs with PHA-stimulatedHIV-1 negative donor PBLs was used for RT-PCR. Phylogenetic analy-sis of sequences derived from cocultivation revealed that this patient wasinfected with subtype B.

e Not done.f Not enough PCR material for sequencing.

226 C. Yang et al. / Journal of Clinical Virology 30 (2004) 224–228

Fig. 1. Phylogenetic classification of the newly sequenced (bold and underlined) sequences inpol protease (A) andenvgp41 (B) regions. The trees werederived from the nucleotide sequence alignments ofprt and gp41 regions (consensus sequences are 252 bp forprt and 378 bp for gp41). Numbers at thenodes indicate the percentage of bootstrap values of 500 replicas in which the cluster is supported; only the values equal to or greater than 75% are shown.

terminator cycle-sequencing ready reaction kit followingthe manufacturer’s protocol. The sequencing reactions werethen run in a 377 DNA sequencer (PE Applied Biosystems,Foster City, CA, USA). Sequencing was done in both di-rections. These sequences, along with reference sequencesfrom the HIV sequence database (Car et al., 1998), werealigned using CLUSTALW (1.74) multiple-sequence align-ment program (Thompson et al., 1994). The genetic dis-tances between strains were computed using DNADIST andphylogenetic trees were constructed after all gaps had been

trimmed using the neighbor-joining method included in thePhylip 3.5c package (Felsenstein, 1993).

3. Results

3.1. HIV serologic tests and viral load measurement

To confirm the HIV-1 serologic status of the four US pa-tients (no plasma was available from patient material sent

C. Yang et al. / Journal of Clinical Virology 30 (2004) 224–228 227

from New Zealand), an aliquot of the plasma samples wasre-tested for anti-HIV antibodies at the CDC HIV diagnos-tics reference laboratory. All patients were strongly reac-tive on HIV-1/2 EIA and were confirmed by HIV-1 West-ern blot assay to have HIV-1 antibodies. All four patientshad undetectable viral load in their plasma samples. In ad-dition, the supernatant of cocultivated patient’s PBLs withPHA-stimulated HIV-1 negative donor PBLs from the pa-tient 5 also had undetectable viral load when UltraSensitiveAmplicor HIV-1 MonitorTM, version 1.5 was used (detec-tion limit <50 copies/ml of plasma) (Table 1).

3.2. Molecular confirmation of HIV-1 infections and viralsequence analysis

Of the four patients whose plasmas were available, three(patients 1, 2, and 3) were positive by RT-PCR in at least oneof the three regions tested (integrase, protease, orenvgp41).The fourth patient (patient 5) had positive RT-PCR signalsin these regions only when RNA isolated from the super-natant of cocultivated patient PBLs with PHA-stimulatedHIV-1 negative donor PBLs was used (Table 1). We nexttested DNA obtained from PBMCs of the patients. Four ofthe five patients had positive signals in at least two of thethree regions tested (integrase, protease, orenvgp41), whilethe fifth patient (patient 5) had positive PCR results in allthree regions only when DNA isolated from PHA-stimulatedpatient’s PBLs was used. All five patients were RT-PCR andDNA-PCR negative for HIV-2 protease.

To classify the HIV-1 strains from the five patients intotheir respective phylogenetic lineages, selected PCR prod-ucts were sequenced and phylogenetic analysis was con-ducted along with HIV-1 reference sequences from the HIVsequence database (Fig. 1). Analysis ofenvgp41 and pro-tease regions revealed that all patients were infected withHIV-1 group M viruses. Analysis ofenvgp41 region in-dicated that three patients (patients 1, 3, and 5) were in-fected with group M subtype B viruses. Patient 4 (from NewZealand, a refugee from Sri Lanka) was infected with a sub-type C virus. Likewise, analysis of thepol protease regionrevealed that patients 1, 3, and 5 were infected with subtypeB viruses while patient 2 (from Guinea) was infected with aCRF 02AG virus. Thus, the subtype identification was con-sistent with the molecular epidemiology of HIV-1 infectionin these diverse geographic areas.

4. Discussion

In the current study, we have used highly sensitive andbroadly reactive generic molecular detection assays (Yanget al., 1999, 2000) to confirm HIV-1 infections from fivepatients for whom viral nucleic acid detection proved to bedifficult by either ultrasensitive viral load assays or in-housePCR detection assays. The use of highly sensitive molec-ular detection tools here allowed us to amplify viral RNA

sequences from three of the four patients for whom plasmasamples were available and proviral DNA sequences fromthe primary PBMCs of all four patients. The fifth patient(patient 5) had positive RT-PCR or PCR results in all threeregions only when RNA isolated from the supernatant ofcocultivated patient PBLs with PHA-stimulated HIV-1 neg-ative donor PBLs or DNA isolated from PHA-stimulatedpatient’s PBLs was used. Phylogenetic analysis indicatedthat three patients from the USA were infected with subtypeB viruses and the fourth patient from the USA who migratedfrom Guinea, was infected with a CRF02AG virus. TheNew Zealand patient was infected with a subtype C virus.These results indicate that viral subtypes may have no effecton the ability of these assays to amplify viral RNA fromplasmas.

It has been reported that Amplicor HIV-1 MonitorTM, ver-sion 1.0 had suboptimal performance when used to quantifynon-subtype B viruses (Emery et al., 2000; Parekh et al.,1999). Therefore, we used the Amplicor HIV-1 MonitorTM,version 1.5 to measure the viral load in the current study,using ultransensitive protocol to increase the lower limit ofdetection. However, all the patients had undetectable viralload (<50 copies/ml). While our assays do not quantify vi-ral copy numbers, we were able to detect viral RNA fromthree of the four patients. The differences in detection be-tween our assays and commercially available ultrasensitiveviral load assays may due to the fact that the primers usedin our assays are located in more conserved regions of theHIV-1 genome than the ones used by commercial ultra-sensitive viral load assays since our previous studies haveshown that these primers could amplify all HIV-1 viral se-quences including groups M, N, and O, while the Amlpi-cor HIV-1 Monitor, version 1.5 can only amplify HIV-1group M viruses (Emery et al., 2000; Müller-Trutwin andBarré-Sinoussi, 1999; Parekh et al., 1999; Yang et al., 1999,2000). Another possible explanation for these differencesis that the patients that we studied here were infected withvariant pol gene regions or had very low viral load sincesamples from patients with lower end of detectibility aresubjected to Poisson distribution, resulting in undetectableviral load by Amplicor ultrasensitive HIV-1 Monitor (v1.5)(Müller-Trutwin and Barré-Sinoussi, 1999). Because of thelimited materials available, we could not further pursuit theseissues. As nucleic acid testing is becoming a standard toolfor monitoring blood safety, generic primers that are capableof amplifying all known HIV-1 subtypes, such as the onesdescribed here, may be considered for future test develop-ment. The adaptation of these primers to develop a real-timequantitative-PCR assay is in progress.

In summary, we have used the highly sensitive andbroadly reactive generic molecular detection assays to con-firm HIV-1 infections from five patients for whom viralnucleic acid detection had proved to be difficult. Taken to-gether, these data provide further evidence that these genericprimers may be excellent candidates for the developmentof clinical nucleic acid-based diagnostic tools.

228 C. Yang et al. / Journal of Clinical Virology 30 (2004) 224–228

References

Car JK, Foley BT, Leitner T, Salminen M, Korber B, McCutchan F.Reference sequences representing the principal genetic diversity ofHIV-1 in the pandemic. In: Korber B, Kuiken C, Foley B, HahnB, McCutchan F, Mellors J, Sodroski J, editors. Human retrovirusesand AIDS—1998. Los Alamos, New Mexico: Los Alamos NationalLaboratory, 1998. p. III10–9.

Emery S, Bodrug S, Richardson BA, Giachetti C, Bott MA, Panteleeff D.Evaluation of performance of the Gen-probe human immunodeficiencyvirus type 1 viral load assay using primary subtype A, C, and Disolates from Kenya. J Clin Microbiol 2000;38:2688–95.

Felsenstein J. PHYLIP (Phylogeny Inference Package), version 3.5c, Dis-tributed by the author. Seattle: Department of Genetics, University ofWashington, 1993.

Hu D, Dondero TJ, Mastro TD, Gayle HD. Global and molecular epidemi-ology of HIV. In: Wormser GP, editor. AIDS and other manifestationof HIV infection. Philadelphia: Lippincott-Raven. 1998. p. 27–40.

Janini LM, Pieniazek D, Peralta JM, Schechter M, Tanuri A, VicenteAC, et al. Identification of single and dual infections with distinctsubtypes of human immunodeficiency virus type 1 by using restrictionfragment length polymorphism analysis. Virus Genes 1996;13:69–81.

Lucas GM, Chaissson RE, Moore RD. Highly active antiretroviral therapyin a large urban clinic: risk factors for virologic failure and adversedrug reactions. Ann Int Med 1999;131:81–7.

Masciotra S, Yang C, Pieniazek D, Thomas C, Owen SM, McLure H.Detection of simian immunodeficiency virus in diverse species andhuman immunodeficiency virus type 2 using consensus primers withinthe pol region. J Clin Microbiol 2002;40:3167–71.

Migueles SA, Sabbaghian MS, Shupert WL, Bettinotti MP, Marincola FM,Martino L. HLA B*5701 is highly associated with restriction of virusreplication in a subgroup of HIV-infected long term nonprogressors.Proc Natl Acad Sci USA 2000;97:2709–14.

Müller-Trutwin M, Barré-Sinoussi F. Towards improvements in molec-ular tools for diagnosis and managemene of HIV infections. Lancet1999;354:1660–2.

Parekh B, Phillips S, Granade TC, Baggs J, Hu DJ, Respess R. Impact ofHIV type 1 subtype variation on viral RNA quantitation. AIDS ResHum Retroviruses 1999;15:133–42.

Propato A, Schiaffella E, Vicenzi E, Francavilla V, Baloni L, ParoliM. Spreading of HIV-specific CD8+ T-cell repertoire in long-termnonprogressors and its role in the control of viral load and diseaseactivity. Hum Immunol 2001;62:561–76.

Sullivan PS, Do A, Ellenberger E, Pau CP, Paul P, Robbins K. Humanimmunodeficiency virus (HIV) subtype surveillance of African-bornpersons at risk for group O and group N HIV infections in the UnitedStates. J Infect Dis 2000;181:463–9.

Thompson JD, Higgins DG, Gibson TJ. CLUSTALW: improving the sen-sitivity of progressive multiple sequence alignment through sequenceweighting. Nucleic Acids Res 1994;22:4673–80.

Yang C, Pieniazek D, Owen SM, Fridlund C, Nkengasong J, MastroTD. Detection of phylogenetically diverse human immunodeficiencyvirus type 1 groups M and O from plasma using highly sensitive andspecific generic primers. J Clin Microbiol 1999;37:2581–6.

Yang C, Dash BC, Simon F, van der Groen G, Pieniazek D, Gao F.Detection of diverse variants of HIV-1 groups M, N, and O and simianimmunodeficiency viruses from chimpanzees using genericpol andenv primer pairs. J Infect Dis 2000;181:1791–5.