AIDS RESEARCH AND HUMAN RETROVIRUSESVolume 20, Number 9, 2004, pp. 1015–1018© Mary Ann Liebert, Inc.

Sequence Note

Construction of an Infectious HIV Type 1 Molecular Clone from an African Patient with a Subtype D/C

Recombinant Virus

BINSHAN SHI,1 SEAN M. PHILPOTT,1 BARBARA WEISER,1,2 CARLA KUIKEN,3 CHERYL BRUNNER,1

GUOWEI FANG,1 KEITH R. FOWKE,4 FRANCIS A. PLUMMER,4 SARAH ROWLAND-JONES,5

JOB BWAYO,6 AGGREY O. ANZALA,6 JOSHUA KIMANI,6 RUPERT KAUL,5

and HAROLD BURGER1,2

ABSTRACT

The majority of HIV-1 infections worldwide occur in Africa, where subtype B viruses are rare and intersub-type recombinants are common. Pathogenesis and vaccine studies need to focus on viruses derived from Africanpatients, and infectious HIV-1 molecular clones can be useful tools. To clone non-B subtypes and recombi-nant viruses from patients, we cultivated HIV-1 from the plasma of a Kenyan long-term survivor. Viral DNAwas cloned into a plasmid, which was transfected into COS cells; progeny virus was propagated in PBMCs.Sequence analyses revealed that both the patient’s plasma HIV-1 RNA and the cloned DNA genomes wererecombinants between subtypes D and C; subtype C sequences comprised the nef and LTR regions. The clonedvirus used the CCR5 coreceptor and did not form syncytia in vitro. This infectious HIV-1 subtype D/C re-combinant molecular clone obtained from a Kenyan long-term survivor promises to be useful to study patho-genesis and vaccine design.

1015

IN THE GLOBAL HIV-1 EPIDEMIC approximately 70% of infec-tions occur in Sub-Saharan Africa, where subtype B virus is

rare.1 In Kenya, multiple non-B HIV-1 subtypes are prevalentincluding A, C, and D2; intersubtype recombinants are alsocommon.2–6 The fields of HIV-1 pathogenesis and vaccine de-sign would benefit from studies of viruses obtained from sin-gle, well-characterized African patients. Infectious molecularclones are useful for such investigation. A limited number ofreplication-competent molecular clones have been derived previously from individuals infected with non-subtype Bstrains.3,7–9 Infectious molecular clones have also been obtainedfrom circulating recombinant forms (CRFs) of HIV-1.10–12 Re-combination provides a mechanism to increase viral sequence

diversity rapidly and may complicate efforts to develop vac-cines. It would be worthwhile to obtain an infectious molecu-lar clone from an African patient infected with a unique inter-subtype HIV-1 recombinant.

We studied HIV-1 isolated from seven women enrolled inthe Pumwani Sex Worker Cohort in Nairobi, Kenya.5,6 The re-search program was approved by the Institutional ReviewBoards at the Kenyatta National Hospital, University of Mani-toba, and New York State Department of Health. The womenwere long-term survivors who had not received antiretroviraltherapy.

We determined complete plasma-derived HIV-1 RNA se-quences from the seven subjects.5,6 One of them, subject

1Wadsworth Center, New York State Department of Health, Albany, New York 12208.2Albany Medical College Albany, New York 12208.3Los Alamos National Laboratory, Los Alamos, New Mexico 87545.4University of Manitoba, Winnipeg, Manitoba, Canada R3EOW3.5Oxford University, Oxford OX3 9DS, UK.6University of Nairobi, Nairobi, Kenya.

5201_a13_p1015-1018 9/16/04 9:15 AM Page 1015

SHI ET AL.1016

FIG. 1. Construction of a full-length molecular clone of HIV-1. (A) Normal donor PBMCs were inoculated with plasma frompatient ML249; HIV-1 was cultivated and high-molecular-weight DNA extracted. PCR primers were designed based on the plasmaviral RNA sequence.6 Long PCR was used to amplify both the 59 end and the 39 end that together cover the full length of theHIV-1 genome. Primer pair for 59 fragment:

Forward FU3: 59TGG AAG GGT TAA TTT ACT CTA AGA AAA GGC 39Reverse R4845: 59GAT CTC TGC TGT CCC TGT AA 39

Primer pair for 39 end fragment

Forward F4077: 59 AGG TCT ACC TAT CAT GGG TAC 39Reverse RU5-EcoRI: 59 CAG ATC TAG GAA TTC TCC ACA CTG ACT AAA GAG GTC 39 (The EcoRI site is under-

lined).

(B) Map of the complete HIV-1 genome of clone pWCML249, a recombinant of subtypes D and C, with clade C sequences inthe nef gene and LTR.

5201_a13_p1015-1018 9/16/04 9:15 AM Page 1016

ML249, was found to have a subtype D and C recombinantstrain, with clade C sequences comprising a large portion of nefand the long terminal repeat (LTR)6 (GenBank accession num-ber AY322187). HIV-1 derived from the patient’s plasma wascultivated in normal donor peripheral blood mononuclear cells(PBMC) by using standard methods.13 To obtain a molecularclone of the HIV-1 strain, high-molecular-weight DNA frominfected donor PBMCs was extracted and used as a templatefor amplification. Polymerase chain reaction (PCR) primerswere designed based on the plasma HIV-1 RNA sequence.6

Long PCR employing the GeneAmp XL PCR Kit (PerkinElmer)was used to amplify two fragments: the 59 end and 39 end thattogether encompass the full length of the HIV-1 genome5,6

(Fig. 1A). First, the 59 end fragment was ligated into the vec-tor pPCR-Script (Stratagene, La Jolla, CA) and a recombinantclone with the correct orientation, pWCML249-5E, was ob-tained. Then, the 39 end fragment was inserted into pW-CML249-53 (Fig. 1A). A full-length clone, pWCML249, wasconstructed and amplified in XL10-Gold Ultracompetent Bac-terial Cells (Stratagene) grown at 25°C in LB broth containing100 mg/ml ampicillin.

The DNA sequence of the clone was determined, and HIV-1 subtypes were identified by using phylogenetic analysis, dis-tance plotting, and bootscanning, the latter two with the helpof SimPlot.14 The complete molecular DNA clone of HIV-1had a sequence of 9721 bp, which was 96% identical to the full-length HIV-1 RNA sequence obtained directly from patientML249’s plasma.6 Both the cloned DNA and parental plasmavirus RNA sequences were found to be HIV-1 recombinantscomposed of subtypes D and C, with C sequences comprisingmost of nef and the U3 region of the LTR (Fig. 1B). A previ-ous study of the complete plasma HIV-1 genomic sequence ob-tained from patient ML249 revealed no obvious attenuating se-quences in nef or the LTR.6

To determine if pWCML249 was infectious, the plasmidpWCML249 was transfected into COS cells by using FuGENE6 (Roche). Three days later, HIV-1 virions were detected in thetransfection culture supernatant by p24 enzyme-linked im-munosorbent assay (ELISA) (NEN Life Sciences Products) andreverse transcriptase assay (Roche). HIV-1 detected in the COSsupernatant then served as inoculum to infect 5 3 106 normaldonor PBMCs. After 24 hr, the cells were washed three timesin phosphate buffered saline to remove the input virus. Virusproduction was assayed at 14 days by p24 antigen assay andconfirmed by RT-PCR.13 Infectious virus obtained at day 14was titered in normal donor PBMCs by standard limiting dilu-tion cultivation,13 and had a titer of 40,000 tissue culture in-fectious doses (TCID50) per milliliter.

After infecting PBMCs with cloned virus, we extracted vi-ral RNA from the culture supernatant and treated it with DNaseI. An 811-bp fragment of env (nt 6338–7149, including regionsV1–V2 of gp120) was amplified by RT-PCR and a 637-bp por-tion was examined by consensus sequencing. The sequence wasfound to match the corresponding sequence of the plasmidclone, indicating that the virus cultivated in PBMCs was de-rived from the molecular clone. To determine the coreceptorusage of virions produced by clone WCML249, we infected aGHOST-CD4 cell system, with cells expressing either CCR5or CXCR4.15 Like most non-subtype B HIV-1 strains isolatedfrom Africans, the clone used coreceptor CCR56 (Fig. 2). Con-sistent with CCR5 usage, cloned virus did not form syncytia inan MT2 cell assay (data not shown).15

The replication kinetics of the infectious molecular clonewere studied in PBMC culture, with HIV-1SF162, an HIV-1laboratory strain that uses CCR5, serving as a control (Fig. 3).The clone grew at a slower pace and to a lower titer in PBMCsthan the laboratory strain, as might be expected for a primary

INFECTIOUS HIV-1 RECOMBINANT MOLECULAR CLONE 1017

FIG. 2. Coreceptor utilization by molecularly cloned virusfrom ML249 as measured by p24 production by virus-infectedGHOST cell cultures at day 10. To determine coreceptor us-age, GHOST-CCR5 (Hi5) and GHOST-CXCR4 cells were in-oculated with a standard quantity of titered cloned virus: 1 3104 tissue culture infective dose (TCID50). All assays were per-formed in triplicate. HIV JR-CSF, LAV, and RF were inocu-lated in parallel as R5-, X4-, and dual-tropic control viruses,respectively. To eliminate any potential artifacts resulting frominfection due to low levels of endogenous coreceptor expres-sion, parental GHOST cells were also inoculated in triplicatewith WCML249 and control isolates.

FIG. 3. Replicative properties of virus from the infectiousmolecular clone WCML249. One and a half million normaldonor PBMCs were infected in quadruplicate with 1500TCID50 of WCML249 virus using standard methods. For com-parison, parallel cultures were inoculated with SF162, a CCR5-using laboratory strain of HIV-1. After 2 hr incubation at 37°C,the cells were washed three times with PBS and resuspendedin 1 ml fresh media. Supernatants were harvested at 0, 1, 2, 3,4, 7, and 10 days after infection and analyzed for HIV-1 p24antigen.

5201_a13_p1015-1018 9/16/04 9:15 AM Page 1017

HIV-1 strain using CCR5 and isolated from a long-term sur-vivor.

This HIV-1 clone is representative of an African primary iso-late. It was obtained from a well-characterized Kenyan long-term survivor who had not received antiretroviral therapy.6 Likemany African patients, she was infected with an intersubtyperecombinant strain. Phylogenetic analysis of the subtype D gagfragment and the subtype C nef sequence revealed that theywere closely related to other East African HIV-1 strains. Fewinfectious molecular clones have been obtained from primaryisolates representing strains in the epidemic in Africa. Thisclone promises to be a useful tool for studies of HIV-1 diseaseprogression and vaccine design.

SEQUENCE DATA

The GenBank accession number for the sequence of the in-fectious HIV-1 molecular clone pWCML249 is AY445524.

ACKNOWLEDGMENTS

We thank the Wadsworth Center Molecular Genetics CoreLaboratory for oligonucleotide synthesis and DNA sequenceanalysis, the NIH AIDS Reference and Reagent Program, andSuzanne Beck for manuscript preparation. This study wasfunded in part by Grant RO1-AI-42555 from the U.S. NationalInstitute of Allergy and Infectious Diseases and by the UK Med-ical Research Council.

REFERENCES

1. Joint United Nations Programme on HIV/AIDS (UNAIDS), WorldHealth Organization (WHO): AIDS Epidemic Update: December,2002. UNAIDS, Geneva.

2. Dowling WE, Kim B, Mason CJ, et al.: Forty-one near full-lengthHIV-1 sequences from Kenya reveal an epidemic of subtype A andA-containing recombinants. AIDS 2002;16:1809–1820.

3. Kuiken C, Foley B, Freed E, et al. (eds.): HIV Sequence Com-pendium 2002. Los Alamos National Laboratory, Los Alamos, NM2002.

4. Najera R, Delgado E, Perez-Alvarez L, and Thomas MM: Genetic

recombination and its role in the development of the HIV-1 pan-demic. AIDS 2002;16:S3–S16.

5. Fang G, Weiser B, Kuiken C, et al.: Recombination following su-perinfection by HIV-1. AIDS 2004;18:153–159.

6. Fang G, Kuiken C, Weiser B, et al.: Long-term survivors inNairobi: Complete HIV-1 RNA sequences and immunogenetic as-sociations. J Infect Dis 2004;190:697–701.

7. Novelli P, Vella C, Oxford J, and Daniels RS: Construction andcharacterization of a full-length HIV-1(92UG001) subtype D in-fectious molecular clone. AIDS Res Hum Retroviruses2002;18:85–88.

8. Ndung’u T, Renjifo B, and Essex M: Construction and analysis ofan infectious human immunodeficiency virus type 1 subtype C mo-lecular clone. J Virol 2001;75:4964–4972.

9. Alizon M, Wain-Hobson S, Montagnier L, et al.: Genetic variabilityof the AIDS virus: Nucleotide sequence analysis of two isolatesfrom African patients. Cell 1986;46:63–74.

10. Kusagawa S, Sato H, Tomita Y, et al.: Isolation and characteriza-tion of replication-competent molecular DNA clones of HIV type1 CRF01 AE with different coreceptor usages. AIDS Res HumRetroviruses 2002;18:115–122.

11. Takahoko M, Tobiume M, Ishikawa K, et al.: Infectious DNA cloneof HIV type 1 A/G recombinant (CRF02 AG) replicable in pe-ripheral blood mononuclear cells. AIDS Res Hum Retroviruses2001;17:1083–1087.

12. Salminen MO, Ehrenberg PK, Mascola JR, et al.: Construction andbiological characterization of infectious molecular clones of HIV-1 subtypes B and E (CRF01 AE) generated by the polymerasechain reaction. Virology 2000;278:103–110.

13. Fang G, Burger H, Grimson R, et al.: Maternal plasma HIV-1 RNAlevel: A determinant and projected threshold for mother-to-childtransmission. Proc Natl Acad Sci USA 1995;92:12100–12104.

14. Lole KS, Bollinger RC, Paranjape RS, et al.: Full-length humanimmunodeficiency virus type 1 genomes from subtype C-infectedseroconverters in India, with evidence of intersubtype recombina-tion. J Virol 1999;73:152–160.

15. Philpott S, Weiser B, Anastos K, et al.: Preferential suppression ofCXCR4-specific strains of HIV-1 by antiviral therapy. J Clin In-vest 2001;107:431–438.

Address reprint requests to:Harold Burger

Wadsworth Center120 New Scotland AvenueAlbany, New York 12208

E-mail: burger@Wadsworth.org

SHI ET AL.1018

5201_a13_p1015-1018 9/16/04 9:15 AM Page 1018