Assessment of the Role of the Central DNA Flap in Human Immunodeficiency Virus Type 1 Replication by Using a Single-Cycle Replication System

In this study, reverse transcriptase (RT)- and integrase (IN)-defectivehuman immunodeficiency virus type 1 (HIV-1) was transcomplementedwith Vpr-RT-IN fusion proteins to delineate pol sequences importantfor HIV-1 replication. Our results reveal that a 194-bp sequenceencompassing the 3'end of the IN gene and containing the centralDNA flap is necessary and sufficient for efficient HIV-1 single-cyclereplication in dividing and nondividing cells. Furthermore,we show that the central DNA flap enhances HIV-1 single-roundreplication by five- to sevenfold, primarily by facilitatingnuclear import of proviral DNA. In agreement with previous reports,our data support a functional role of the central DNA flap duringthe early stages of HIV-1 infection.

INTRODUCTION

Top

Introduction

The complexity of human immunodeficiency virus type 1 (HIV-1)replication is attributed in large part to the intricate interplaythat takes place between cis-acting sequences present on viralnucleic acids and viral or host cell proteins that functionin trans (22). The HIV-1 pol gene encodes three enzymatic proteins,including protease, reverse transcriptase (RT), and integrase(IN), which play critical roles during specific stages of thevirus infection cycle. Soon after virus entry, RT catalyzesthe conversion of the viral RNA genome into double-strandedproviral DNA, while IN mediates proviral DNA integration intothe host cell genome (for a review, see references 10, 15, and21). Even though extensive in vitro biochemical and mechanisticstudies have greatly contributed to a better understanding ofthe primary function and mode of action of RT and IN enzymes,studies performed in the context of HIV-1 infectious proviralclones have revealed that some mutations and/or internal deletionsin RT or IN can significantly alter steps other than reversetranscription or integration, such as virus assembly or release,or even inactivate virus infectivity (1, 2, 4, 14, 20, 35).These pleiotropic phenotypes resulting from mutagenic analysissuggest that RT and IN may play other roles which are independentof their enzymatic activities. Furthermore, introduction ofmutations in RT and/or IN sequences may simultaneously affecta cis-acting element(s) present within the pol gene sequencethat is required for efficient virus replication.

In contrast to oncoretroviruses, HIV-1 and other lentiviruseshave the capacity to infect nondividing cell populations, suchas macrophages, mucosal dendritic cells, and nondividing T cells,since they do not depend on host cell mitosis to mediate thenuclear translocation of their preintegration complex (PIC)(5, 12, 27, 28, 31, 40). At the molecular level, HIV-1 PIC nucleartransport was shown to proceed through intact nuclear pore complexesby an active and energy-dependent mechanism that involves thekaryophilic properties of several PIC-associated viral proteins,including Matrix (MAp17^gag), IN, and Vpr (3, 6, 12, 19, 20,23, 24, 32, 39). In addition, a cis-acting element designatedthe central DNA flap located in the 3' region of the pol genesequence was also shown to contribute to the nuclear importof HIV-1 proviral DNA in both dividing and nondividing cells(45). The central DNA flap is a region of triple-stranded DNAcreated by two discrete half-genomic fragments with a centralstrand displacement event controlled in cis by a central polypurinetract (cPPT) and a central termination sequence (CTS) duringHIV-1 reverse transcription (7, 8). HIV-1 viruses carrying aninactivated cPPT or CTS were reported to exhibit a considerableimpairment of viral replication in different dividing and nondividingtarget cells (7, 8), presumably because of a defect at the levelof the nuclear import of the PIC (45). However, these resultshave been put into question recently by two other studies (13,29), which provided evidence indicating that the central DNAflap did not play a major role in either PIC nuclear importor HIV-1 replication in a variety of cell lines. Interestingly,in contrast to these studies that used replication-competentviruses, numerous other studies have reported that the centralDNA flap conferred a transduction advantage of approximately2- to 10-fold on HIV-1-derived lentiviral vectors, thus suggestingthat the central DNA flap facilitated an early step(s) in lentiviralinfection (11, 17, 30, 33, 34, 37, 46). The exact impact ofthe central DNA flap on the early steps of HIV-1 infection isstill an open question that remains to be clarified.

Requirement of pol gene sequence for efficient HIV-1 single-cycle infection.

Top

Requirement of pol gene...

In order to investigate the impact of the HIV-1 pol gene sequenceon virus replication, we generated a previously described HIV-1RT and IN transcomplemented replication system (16, 41). AnRT- and IN-defective HIV-1 provirus (R^-/RI^-) with an intactpol gene sequence was constructed by replacing the first twoamino acids of RT with two premature stop codons (TGA TAG) ina Vpr- and Nef-defective HxBc2-derived HIV-1 provirus (R^-) byusing a two-step PCR-based method (44). To transcomplement theRT and IN defect of HIV-1 virus, a Vpr-RT-IN fusion proteinexpression plasmid (CMV-R-RT-IN) was also made by insertinga PCR-amplified HIV-1 RT and IN gene cDNA in frame with Vprinto the SVCMV-Vpr plasmid (44) (Fig. 1A). It has been shownthat HIV-1 RT and IN enzymatic defects can be restored in transthrough Vpr-mediated virion incorporation of Vpr-RT-IN fusionprotein, leading to production of viral particles that can undergoa single-round replication (16, 41). To test the infectivityof the transcomplemented RT- and IN-defective virus, 293T cellswere transfected with R^-/RI^- provirus or cotransfected withCMV-R-RT-IN plasmid by using the calcium phosphate DNA precipitationmethod (42). In parallel, the wild-type (wt) provirus (R^-) wasused as a positive control. At 48-h posttransfection, virusstocks were generated from supernatants and quantitated by p24measurements by using an HIV-1 p24 enzyme-linked immunosorbentassay (ELISA) kit (AIDS Vaccine Program of the Frederick CancerResearch and Development Center) as described previously (43).Then the infectivity of each virus was examined by infectingHeLa-CD4-ß-galactosidase (HeLa-CD4-[ß-Gal])cells with equal amounts (15 ng of p24^gag antigen/well) of virusand evaluated by MAGI assay 48-h postinfection (p.i.), as describedpreviously (26). Consistent with a previous report (41), RT-and IN-defective R^-/RI^- viruses were found to be infectiousby MAGI assay only when they were transcomplemented with RTand IN during viral production, reaching infectivity levelscorresponding to 40 to 50% of the wt level (R^- virus) (Fig.1B, left panel). To test the efficiency of Vpr-mediated RT andIN transincorporation, transfected 293T cells were radiolabeledand the resulting viral particles were analyzed for viral proteincontent by immunoprecipitation by using anti-HIV-1 serum (42).Results reveal that, as expected, the R^-/RI^- viral particlesdid not contain any RT or IN proteins (Fig. 1C, lane 3), whereasthe transcomplemented R^-/RI^- virus incorporated RT and IN proteinsat levels comparable to those for the wt R^- virus (Fig. 1C,compare lane 4 to lane 2). Interestingly, it was noted thatsignificant amounts of unprocessed Pr55^gag accumulated in R^-/RI^-viral particles (Fig. 1C, compare lanes 3 and 4 to lane 2).This maturation defect is likely to result from an impairmentof protease activation, given that the Gag-Pol polyprotein precursorwas truncated by early termination of RT and IN and is believedto be responsible for the observed 50 to 60% reduction of virusinfectivity (Fig. 1B).

View larger version (39K):
[in this window]
[in a new window]
FIG. 1. Effect of the HIV-1 IN and RT gene sequences on the infectivity of RT/IN-transcomplemented virus. (A) Schematic structure of HIV-1 proviruses carrying a mutation and/or deletions in pol and of the plasmid encoding the Vpr-RT-IN fusion protein. Provirus R^-/RI^- was constructed by replacing the first two amino acids of RT with two premature stop codons (TGA TAG) in HxBruR^-(R^-) provirus. In R^-/RI^-/ ${Delta}$ IN/flap^- provirus, a 610-bp fragment of the IN gene sequence (including cPPT/CTS) was deleted. The R^-/RI^-/ ${Delta}$ IN/flap⁺, R^-/RI-861, R^-/RI-1798, and R^-/ ${Delta}$ RI proviruses harbor different deletions within the RT and/or IN gene sequences but contain the 194-bp sequence in the 3'-end region of IN, which harbors the cPPT/CTS cis-acting elements. PR, protease. (B) The infectivity of trans-complemented virus produced in 293T cells was evaluated by MAGI assay. Equal amounts (15 ng of p24^gag antigen) of the different viruses were used to infect HeLa-CD4-ß-Gal cells, and the number of infected cells was monitored by X-Gal (5-bromo-4-chloro-3-indolyl-ß-D-galactopyranoside) staining. The infectivity (% infectivity) of each virus stock was calculated as the ratio of the number of ß-Gal-positive cells relative to the number of ß-Gal-positive cells obtained with the wt virus (R^-). The number of ß-Gal-positive cells detected with the R^- virus ranged between 330 to 410 and was set at 100%. The results are representative of three independent experiments. (C) To evaluate Vpr-mediated transincorporation of RT and IN in viral particles, radiolabeled viruses were isolated from cell supernatants, lysed, immunoprecipitated with anti-HIV antibodies, and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (12.5% acrylamide). LTR, long terminal repeat.

We next analyzed the requirement of RT and IN gene sequencesfor virus replication by using this Vpr-RT-IN transcomplementationsystem. A series of proviruses with deletions of the RT and/orIN gene derived from the R^-/RI^- provirus were constructed asindicated in Fig. 1A. In the R^-/RI^-/ ${Delta}$ IN/flap^- provirus, a 610-bpsequence encompassing a large part of the IN gene and includingthe cPPT and CTS cis-acting sequences (from nucleotide 3,912to 4,522; +1 corresponds to the transcription initiation siteof the BRU strain), which were previously shown to play an importantrole in HIV-1 replication, were deleted (7, 8, 11). In the R^-/RI^-/ ${Delta}$ IN/flap⁺provirus, a smaller deletion of 494 bp was introduced, thusleaving intact a 194-bp sequence, including cPPT/CTS elementsat the 3' end of the IN gene. To further test the impact ofthe RT gene sequence on virus replication, different regionsof the RT gene sequence were further deleted, based on the R^-/RI^-/ ${Delta}$ IN/flap⁺provirus and designated R^-/RI-861, R^-/RI-1798, and R^-/ ${Delta}$ RI (Fig.1A). In the R^-/ ${Delta}$ RI provirus, all the RT and IN gene sequencesexcept the 194 bp containing the cPPT/CTS sequences were deleted(a deletion of 2,193 bp). The infectivity of each transcomplementedvirus with a deletion of the RT and/or IN gene was analyzedby MAGI assay. Deletion of the 3' region of IN encompassingthe cPPT/CTS elements resulted in a substantial five- to sevenfolddecrease of viral infectivity compared to that of the transcomplementedR^-RI^- virus (Fig. 1B, left panel). This sharp decrease in viralinfectivity was not due to variation in the levels of RT and/orIN transincorporated into viral particles, since they were foundto be similar in both transcomplemented viruses (Fig. 1C, comparelanes 7 and 5). In contrast, maintenance of a 194-bp sequencein the 3' region of the IN gene sequence, which includes thecPPT/CTS elements (R^-/RI^-/ ${Delta}$ IN/flap⁺), restored infectivity toa level similar to that of the transcomplemented R^-/RI^- virus(Fig. 1B, left panel). Interestingly, deletion of RT gene sequences(R^-/RI-861, R^-/RI-1798, and R^-/ ${Delta}$ RI) had no impact on viral infectivityas long as the 194-bp sequence in the 3' end region of IN wasintact (Fig. 1B, right panel). Furthermore, maintenance of this194-bp fragment was found to confer a six- to sevenfold infectivityadvantage to single-cycle replicating virus in dividing andaphidicolin growth-arrested C8166 T cells (data not shown).Overall, these results indicate that the 3' region of the INgene sequence harbors cis-acting determinants(s) that substantiallyenhance the replication of HIV-1 toward dividing and nondividingCD4⁺ T cells in the context of a single-cycle infection system,while other RT and IN gene sequences are clearly dispensable.

The cPPT contributes to efficient HIV-1 single-cycle replication.

Top

The cppt contributes to...

To further confirm that cis-acting element(s) in the 3' regionof the IN gene sequence contribute to efficient single-cyclereplication, we introduced a 10-bp substitution in the cPPTelement in the R^-/ ${Delta}$ RI provirus and generated a cPPT-defectivemutant designated R^-/ ${Delta}$ RI/cPPT^- (Fig. 2A). These specific mutationsin the cPPT element have previously been reported to preventthe formation of the central DNA flap during reverse transcription(13, 29, 45). Prior to testing virus infectivity, the levelsof transincorporated RT and IN in both the cPPT mutant and thecontrol virus were examined by radiolabeling and immunoprecipitationas described in Fig. 1. Similar levels of virion-associatedRT, IN, p24^gag, and p55^gag were detected in transcomplementedR^-/RI^-, R^-/ ${Delta}$ RI, and R^-/ ${Delta}$ RI/cPPT^- virus preparations (Fig. 2B).To compare the replication potential of transcomplemented R^-/ ${Delta}$ RI/cPPT^-and R^-/ ${Delta}$ RI viruses, CD4⁺ MT4 cells and phytohemagglutinin-stimulatedhuman peripheral blood mononuclear cells (h-PBMCs) were infectedwith equal amounts of each virus stock for 8 h, and at differenttime intervals, virion-associated p24^gag antigen levels in thesupernatant were measured by anti-p24 ELISA. Disruption of thecPPT was found to decrease by five- to sevenfold viral replicationin both MT4 T cells and activated h-PBMCs compared to that inthe trans-complemented R^-/ ${Delta}$ RI control virus (Fig. 2C and D).Hence, we conclude that the central DNA flap is the necessarydeterminant in the 3' region of the IN gene sequence that contributesto efficient single-cycle virus replication.

View larger version (35K):
[in this window]
[in a new window]
FIG. 2. The central DNA flap contributes to efficient HIV-1 single-cycle replication. (A) Schematic structure of HIV-1 provirus with the RT/IN gene deleted (R^-/ ${Delta}$ RI) and the cPPT^- mutant (R^-/ ${Delta}$ RI/cPPT^-). The cPPT element was inactivated by introduction of ten nucleotide substitution mutations, as indicated. (B) To evaluate RT and IN transincorporation, [³⁵S]-methionine-radiolabeled viruses were collected from transfected 293T cells, lysed, and analyzed by immunoprecipitation by using anti-HIV antibodies. To test the replication potential of each virus stock, CD4⁺ MT4 T cells (C) or PHA-stimulated human PBMCs (D) were infected with equal amounts of R^-/ ${Delta}$ RI or R^-/ ${Delta}$ RI/cPPT^- viruses. At different time intervals after infection, viral production was monitored in the supernatants by measurement of HIV-1 p24^gag antigen by using a p24 ELISA assay. The results are representative of two independent experiments.

Effect of the central DNA flap on late-reverse-transcribed DNA products, viral cDNA nuclear import, and proviral DNA integration in h-PBMCs.

Top

Effect of the central...

To investigate the mechanism(s) underlying the action of thecentral DNA flap during single-cycle replication, we first analyzedthe efficiency of proviral DNA integration in h-PBMCs infectedwith cPPT⁺ or cPPT^- transcomplemented viruses by using a previouslydescribed, sensitive, two-step Alu-PCR technique (9). Resultsreveal that levels of integrated proviral DNA detected in theR^-/ ${Delta}$ RI/cPPT^- sample were five- to sevenfold lower than thosedetected in the R^-/ ${Delta}$ RI sample at both 24- and 36-h p.i. (Fig.3A and B). Interestingly, this significant decrease in proviralDNA integration correlated well with the five- to sevenfoldreplication defect observed with the cPPT-defective virus, indicatingthat the central DNA flap contributes to efficient viral single-cyclereplication by acting on an early stage(s) of viral replicationat and/or prior to viral integration. To further investigateat which early step(s) of the infection cycle the central DNAflap acts, we analyzed the total amounts of viral cDNA presentat different early time points by PCR following infection ofh-PBMCs with equal amounts of R^-/ ${Delta}$ RI or R^-/ ${Delta}$ RI/cPPT^- viruses.Briefly, an equal number (2 x 10⁶ cells) of h-PBMCs were lysedin lysis buffer (20 mM Tris-HCl [pH 8.0], 100 mM KCl, 0.05%NP-40, 0.05% Tween 20) and treated with proteinase K (100 µg/ml)prior to phenol-chloroform DNA purification. Then fivefold-seriallydiluted DNA samples were subjected to PCR analysis with specificprimers (5'-U3, 5'-GGATGGTGCTTCAAGCTAGTACC-3', and 3'-Gag, 5'-ACTGACGCTCTCGCACCCATCTCTCTC-3')and further analyzed by Southern blotting by using specificPCR DIG-Labeling probes (Roche Diagnostics, Laval, Quebec, Canada).As shown in Fig. 4A, at 6-h p.i., similar amounts of total viralcDNA were detected in R^-/ ${Delta}$ RI- and R^-/ ${Delta}$ RI/cPPT^--infected cells(4.9 copies/cell versus 4.6 viral copies/cell), suggesting thatboth transcomplemented viruses entered cells with similar efficienciesand underwent uncoating and reverse transcription at comparablerates. In contrast, between 12- and 24-h p.i., total amountsof late cPPT^- reverse-transcribed products decreased at a ratethat was clearly different from that of viral cPPT⁺ cDNA products.At 12-h and 24-h p.i, levels of viral cPPT^- cDNA were reducedby approximately 45 and 40% compared to the levels of cPPT⁺cDNA, which stayed quite stable during the same time interval(between approximately 87 and 88% of their levels at 6 h) (Fig.4A). At later time points (between 24 and 48 h), both cPPT⁺and cPPT^- viral cDNAs decreased at similar rates, most probablyas a result of the dilution of unintegrated viral cDNA thatoccurs upon cell division. This difference in the rate of viralcDNA decrease detected between R^-/ ${Delta}$ RI and R^-/ ${Delta}$ RI/cPPT^- infectionwas not due to intrinsic variation between samples, since similarlevels of a control cellular DNA (human ß-2-adrenergicreceptor [ß2-AR] gene) were detected by PCR in eachsample (Fig. 4B, left panel). These results suggest that thecentral DNA flap does not interfere with the rate of the reversetranscription step per se but appears to influence the rateof accumulation of total viral cDNA product.

View larger version (37K):
[in this window]
[in a new window]
FIG. 3. Effect of the central DNA flap on HIV-1 proviral DNA integration in h-PBMC. (A) h-PBMCs were infected with R^-/ ${Delta}$ RI or R^-/ ${Delta}$ RI/cPPT^- virus (125 ng of p24/10⁶ cells). At 24- and 36-h p.i., the cells were lysed, and serially diluted cell lysates were analyzed by two-step Alu-PCR and Southern blotting for specific detection of integrated proviral DNA from infected PBMCs (upper panel) or from ACH-2 cells as the quantitative control (lower panel). (B) Quantitative analysis of integrated proviral DNA in single-cycle infection. The bands in panel A were quantified by laser densitometry, and the number of integrated proviral DNA copies per cell was determined by using the PCR-generated standard curve derived from ACH-2 cells.

View larger version (33K):
[in this window]
[in a new window]
FIG. 4. Effect of the central DNA flap on HIV-1 cDNA nuclear import. (A) h-PBMCs were infected with transcomplemented R^-/ ${Delta}$ RI (cPPT⁺) and R^-/ ${Delta}$ RI/cPPT^- (cPPT^-) viruses (125 ng of p24/10⁶ cells) for 2 h. As the negative control, 3'-azido-3'-deoxythymidine (AZT; 10 µM)-pretreated PBMCs were infected with the same amounts of transcomplemented R^-/ ${Delta}$ RI virus. At each indicated time point, serial dilutions of extracted total DNA were analyzed for late-reverse-transcription products by PCR by using long terminal repeat (LTR)-Gag primers and Southern blotting. HIV-1 late-reverse-transcription products detected in the left panel were quantified by laser densitometry. The diagram at the right shows the number of HIV-1 cDNA copies per cell as determined by using the PCR-generated standard curve (B, right panel). These results are representative of those obtained in two independent experiments. Serially diluted R-/ ${Delta}$ RI plasmid DNA was used as a standard for DNA copy quantification (right panel). To evaluate cellular DNA levels in each sample, the cellular ß2-AR gene was amplified by PCR and visualized by ethidium bromide staining (left panel). (C) At 24-h p.i., 2 x 10⁶ infected h-PBMCs were fractionated into cytoplasmic and nuclear fractions as described previously (36). The amounts of viral DNA in the cytoplasmic and nuclear fractions were evaluated by PCR by using HIV-1 LTR-Gag primers and were visualized by ethidium bromide staining. The R^-/ ${Delta}$ RI plasmid DNA was used as a PCR-positive control (pc) (upper panel). In parallel, the purity and DNA content of each subcellular fraction were evaluated by PCR detection of the human globin gene and mitochondrial DNA and were visualized by ethidium bromide staining (lower panel). N, nuclear fraction; C, cytoplasmic fraction.

In parallel, we analyzed viral cDNA nuclear import by subcellularfractionation and subsequent detection of viral cDNA associatedwith nuclear or cytoplasmic fractions as previously described(36). Human PBMCs were infected with equivalent amounts of transcomplementedR^-/ ${Delta}$ RI/cPPT^- or R^-/ ${Delta}$ RI viruses, and cytoplasmic and nuclear fractionswere isolated from the same number of cells at 24-h p.i. Allfractions were then analyzed by PCR as described in the legendto Fig. 4A. The presence of total viral DNA was visualized byethidium bromide staining, and the staining intensity of eachamplified DNA product was quantified by using a ChemiImager5500 system with AlphaEaseFC software (Alpha Innotech Corporation).The results shown in Fig. 4C (upper panel) reveal that at 24-hp.i. total amounts of viral cDNA in the R^-/ ${Delta}$ RI-infected sample(including cytoplasmic and nuclear fractions) were approximatelythreefold higher than those detected in the R^-/ ${Delta}$ RI/cPPT^- infectedsample, thus confirming the data obtained from Fig. 4A. Interestingly,while approximately 75% of total viral cDNA was detected inthe nuclear fraction of R^-/ ${Delta}$ RI-infected cells, only 30% of totalviral cDNA was found in the nuclei of R^-/ ${Delta}$ RI/cPPT^--infected cells(Fig. 4C, upper panel). Moreover, the absolute levels of nuclear-associatedviral cDNA were approximately sevenfold higher with the wt virusthan with the cPPT-defective virus. The integrity of the fractionationprocedure was validated by detection of mitochondrial DNA andß-globin DNA, as described previously (18, 36, 38).Results showed that mitochondrial and ß-globin DNAswere found solely in the cytoplasm and the nucleus, respectively(Fig. 4C, lower panel). In addition, levels of each of thesecontrol cellular DNAs were similar in both R^-/ ${Delta}$ RI and R^-/ ${Delta}$ RI/cPPT^-subcellular fractions, confirming that equivalent amounts ofnuclear and cyctoplasmic fractions were analyzed.

Several recent studies have investigated the role of the centralDNA flap in HIV-1 replication and nuclear import and have reachedconflicting conclusions (11, 13, 17, 29, 30, 33, 34, 37, 45,46). In this study, we reexamined this question by using RTand IN transcomplemented HIV-1 viral particles capable of asingle round of replication. Our results reveal that the centralDNA flap was not essential to HIV-1 replication but conferreda five- to sevenfold infectivity advantage to single-cycle replicatingviruses in a variety of cellular systems, including MAGI cells,MT4, dividing and nondividing C8166 T-cell lines, and h-PBMCs(Fig. 1and 2 and data not shown). These results are consistentwith findings reported by several previous studies that thecentral DNA flap conferred a transduction advantage of 2- to10-fold on vesicular stomatitis virus-G-pseudotyped HIV-1 vectors(11, 13, 17, 29, 30, 33, 34, 37, 45, 46). At this point, itis still unclear why this substantial defect in single-roundinfectivity caused by disruption of the DNA flap does not translateinto a detectable difference when the replication kinetic ismonitored by using replication-competent virus, as shown recently(13, 29). Clearly, more studies in this area are required tounderstand this discrepancy.

In an attempt to understand the mechanism(s) underlying theeffect of the central DNA flap during HIV-1 single-cycle replication,we analyzed by PCR the amount of integrated proviral DNA inthe presence or absence of the central DNA flap. Our resultsclearly show that disruption of the DNA flap results in a five-to sevenfold decrease in proviral DNA integration (Fig. 4),suggesting that the central DNA flap contributes to efficientsingle-cycle viral replication by acting on an early stage(s)of the HIV-1 infection cycle at and/or prior to viral integration.We further analyzed total late-reverse-transcribed DNA productsover time during cPPT⁺ or cPPT^- single-cycle viral infectionand determined the distribution of total viral cDNA in the nucleusand the cytoplasm. Our results reveal that the presence of thecentral DNA flap does not significantly influence the amountof viral transcripts produced at early time points (6 h) butcontributes primarily to an accumulation of viral cDNA in thenucleus (Fig. 4). These results are consistent with findingsmade by several previous studies using either replication-competentviruses or single-round HIV-1 vector transduction systems thatthe central cDNA flap enhances the establishment of HIV-1 infectionby facilitating the nuclear import of proviral DNA (17, 30,45).

In addition to the effect on viral cDNA nuclear import, ourdata also suggest that the presence of the central DNA flapmight have a stabilizing and/or a protective effect on viralcDNA (Fig. 4). Indeed, it is possible that the central DNA flapmight contribute to a correct conformation of viral cDNA and/orbe implicated in the recruitment of host cell proteins to forma functional PIC capable of effective proviral DNA nuclear import.Absence of an intact central DNA flap might lead to immaturePIC where viral cDNA is less stable or subject to rapid degradation.In this regard, it has recently been reported that the centralDNA flap region of viral cDNA was resistant to DNase I digestionwhen viral PIC complexes were isolated from the cytoplasm ofinfected cells at 10 h, whereas it was sensitive to degradationwhen complexes were isolated at 8.5 h after infection (25).Alternatively, it may also be possible that the presence ofthe central DNA flap positively modulates the stability of viralcDNA in the nucleus. Interestingly, a similar difference inthe rate of decline of wt and cPPT-defective total viral cDNAwas observed by Limon et al. (29) in infected PBMCs, althoughthe effect was not as pronounced as in our study. Altogether,these observations point toward a possible role of the centralDNA flap in the formation and maturation of HIV-1 PICs; sucha role is likely to influence viral cDNA stability and nuclearimport.

ACKNOWLEDGMENTS

We thank Nicole Rougeau, Johanne Mercier, and Serge Senechalfor technical support. We thank Sylvie Beaulieu, Luchino Y.Cohen, and Ghislaine Duisit for fruitful discussions. We arealso grateful to M. Emerman for the HeLa-CD4-ß-Galcells, which were obtained through the AIDS Research and ReferenceReagent Program, Division of AIDS, NIAID, NIH.

Xiaojian Yao is a recipient of a Médecine-Relève2000-Messenger Foundation Award from the Faculté de Médecine,Université de Montréal. Éric A. Cohen isthe recipient of the Canada Research Chair in Human Retrovirology.This work was supported by grants from the Canadian Foundationfor AIDS Research (CANFAR) (X.Y), the Canadian Institute ofHealth Research (CIHR), and the Fonds de la Recherche en Santédu Québec (FRSQ) (E.A.C.).

FOOTNOTES

* Corresponding author. Mailing address: University of Montreal, Microbiology and Immunology, CP 6128, Station Centre-Ville, Montreal, PQ, H3C 3J7 Canada. Phone, fax, and e-mail for Éric A. Cohen: Phone: (514) 343-5967. Fax: (514) 343-5995. E-mail: eric.cohen{at}umontreal.ca. Phone, fax, and e-mail for Xiaojian Yao: Phone: (514) 343-6280. Fax: (514) 343-5985. E-mail: xiaojian.yao{at}umontreal.ca.

REFERENCES

Top