Production of HIV-1 by resting memory T lymphocytes.

BACKGROUND: The persistence of HIV-1 within resting memory CD4 T cells constitutes a major obstacle in the control of HIV-1 infection. OBJECTIVE: To examine the expression of HIV-1 in resting memory CD4 T cells, using an in-vitro model. DESIGN AND METHODS: Phytohaemagglutinin-activated peripheral blood mononuclear cells were challenged with T cell-tropic and macrophage-tropic HIV-1 clones, and with a replication-incompetent and non-cytotoxic HIV-1-derived vector (HDV) pseudotyped by the vesicular stomatitis virus glycoprotein G. To obtain resting memory CD4 T cells containing HIV-1 provirus, residual CD25(+), CD69(+) and HLA-DR(+) cells were immunodepleted after a 3 week cultivation period. RESULTS: In spite of the resting phenotype, the majority of provirus-harbouring T cells expressed HIV-1 genomes and produced infectious virus into cell-free supernatant. The expression of HDV dropped by only 30% during the return of activated HDV-challenged cells into the quiescent phase. Although resting memory T cells generated in vitro expressed HIV-1 and HDV genome when infected during the course of the preceding T cell activation, they were resistant to HIV-1 and HDV challenge de novo. The infected culture of resting memory T cells showed a higher resistance to the cytotoxic effects of HIV-1 in comparison with the same cultures after reactivation by phytohaemagglutinin. CONCLUSION: The majority of resting memory T cells infected during the course of a preceding cell activation produces virus persistently, without establishing a true HIV-1 latency. The described system could be used as a model for testing new drugs able to control residual HIV-1 replication in resting memory T cells.

The tyrosine kinase Hck is an inhibitor of HIV-1 replication counteracted by the viral vif protein.

The virus infectivity factor (Vif) protein facilitates the replication of human immunodeficiency virus type 1 (HIV-1) in primary lymphocytes and macrophages. Its action is strongly dependent on the cellular environment, and it has been proposed that the Vif protein counteracts cellular activities that would otherwise limit HIV-1 replication. Using a glutathione S-transferase pull-down assay, we identified that Vif binds specifically to the Src homology 3 domain of Hck, a tyrosine kinase from the Src family. The interaction between Vif and the full-length Hck was further assessed by co-precipitation assays in vitro and in human cells. The Vif protein repressed the kinase activity of Hck and was not itself a substrate for Hck phosphorylation. Within one single replication cycle of HIV-1, Hck was able to inhibit the production and the infectivity of vif-deleted virus but not that of wild-type virus. Accordingly, HIV-1 vif- replication was delayed in Jurkat T cell clones stably expressing Hck. Our data demonstrate that Hck controls negatively HIV-1 replication and that this inhibition is suppressed by the expression of Vif. Hck, which is present in monocyte-macrophage cells, represents the first identified cellular inhibitor of HIV-1 replication overcome by Vif.

Characterization of human immunodeficiency virus type 1 vif gene in long-term asymptomatic individuals.

We have determined the sequence of the human immunodeficiency virus type 1 (HIV-1) vif genes from a cohort of 42 long-term nonprogressors (LTNP) and compared these sequences to those of 8 late progressors. The coding potential of the vif open reading frame directly derived by nested PCR from uncultured peripheral blood mononuclear cell DNA was conserved in all 50 individuals. The nucleotide distances between vif sequences were not significantly different between LTNP and late progressors, indicating similar selections of viruses within both types of long-term HIV-1-infected subjects. However, a statistically significant correlation between an amino acid signature at position 132 of Vif and the viral load was found within LTNP. Namely, amino acid Ser was associated with low viral load and amino acid Arg with high viral load. This signature was also observed when LTNP with low viral load were compared to progressors. The Ser132 signature was introduced in place of Arg132 present in the HIV-1 YU-2 Vif prototype into chimeric viruses to assess the impact of Vif signature on the virus. While the replication properties in the SupT1 cell line were unmodified, the mutagenized virus revealed a fivefold decreased replication in activated PBMC, suggesting a possible role of this Vif signature for viral production in vivo.

The HIV-1 Vpr co-activator induces a conformational change in TFIIB.

Vpr is a HIV-1 virion-associated protein which plays a role in viral replication and in transcription and cell proliferation. We have previously reported that Vpr stimulates transcription of genes lacking a common DNA target sequence likely through its ability to interact with TFIIB. However, the molecular mechanism of the Vpr-mediated transcription remains to be precisely defined. In this in vitro study, we show that the binding site of Vpr in TFIIB overlaps the domain of TFIIB which is engaged in the intramolecular bridge between the N- and C-terminus of TFIIB, highly suggesting that binding of Vpr may induce a change in the conformation of TFIIB. Indeed, with a partial proteolysis assay using V8 protease, we demonstrate that Vpr has the ability to change the conformation of TFIIB. We investigated in this partial proteolysis assay a series of Vpr-mutated proteins previously defined for their transactivation properties. Our data show a correlation between the ability of Vpr-mutated proteins to stimulate transcription and their ability to induce a conformational change in TFIIB, indicating a functional relevance of the Vpr-TFIIB interaction.

DNA repair enzyme uracil DNA glycosylase is specifically incorporated into human immunodeficiency virus type 1 viral particles through a Vpr-independent mechanism.

The Vpr protein, encoded by the human immunodeficiency virus type 1 (HIV-1) genome, is one of the nonstructural proteins packaged in large amounts into viral particles. We have previously reported that Vpr associates with the DNA repair enzyme uracil DNA glycosylase (UDG). In this study, we extended these observations by investigating whether UDG is incorporated into virions and whether this incorporation requires the presence of Vpr. Our results, with highly purified viruses, show that UDG is efficiently incorporated either into wild-type virions or into Vpr-deficient HIV-1 virions, indicating that Vpr is not involved in UDG packaging. Using an in vitro protein-protein binding assay, we reveal a direct interaction between the precursor form of UDG and the viral integrase (IN). Finally, we demonstrate that IN-defective viruses fail to incorporate UDG, indicating that IN is required for packaging of UDG into virions.

Priming with tat-deleted caprine arthritis encephalitis virus (CAEV) proviral DNA or live virus protects goats from challenge with pathogenic CAEV.

We previously reported that infection of goats with caprine arthritis encephalitis virus (CAEV) tat- proviral DNA or virus results in persistent infection, since the animals seroconverted and direct virus isolation from cultures of blood-derived macrophages was positive. In this study we wanted to determine whether goats injected with CAEV tat- proviral DNA or virus were protected against challenge with the pathogenic homologous virus and to investigate whether CAEV tat- was still pathogenic. All animals injected with CAEV tat- became infected as indicated by seroconversion and virus isolation. Challenge at 8 or 9 months postinfection demonstrated protection in four of four animals injected with CAEV tat- but did not in three of three mock-inoculated challenged goats. Challenge virus was undetectable in the blood macrophages of protected animals during a period of 6 or 10 months postchallenge. In two of four protected animals, however, we were able to detect the challenge wild-type virus by reverse transcriptase PCR on RNA directly extracted from synovial membrane cells surrounding the inoculation site. This result suggests that protection was achieved without complete sterilizing immunity. Animals injected with CAEV tat- and mock challenged developed inflammatory lesions in the joints, although these lesions were not as severe as those in CAEV wild-type-injected goats. These results confirm the dispensable role of Tat in CAEV replication in vivo for the establishment of infection and pathogenesis and demonstrate in another lentivirus infection model the efficacy of live attenuated viruses to induce resistance to superinfection.

A role for human immunodeficiency virus type 1 Vpr during infection of peripheral blood mononuclear cells.

Studies analysing human immunodeficiency virus type 1 replication in primary cells have demonstrated that Vpr, although dispensable, plays a role along with the matrix (MA) protein in allowing nuclear localization of viral preintegration complexes in non-dividing monocyte-derived macrophages (MDMs). In the current study, experimental infection conditions to analyse the role of Vpr, independently of MA, during infection of PHA/IL-2-stimulated peripheral blood mononuclear cells (PBMC) were designed. It was shown that the absence of Vpr results in a subtle effect on virus production in long-term infection. PCR analysis of the steps of virus retrotranscription during a single cycle of replication in stimulated PBMC revealed that the absence of Vpr alone correlates with an impairment in the nuclear localization of viral DNA. Our data indicate that Vpr is involved in the virus life-cycle during infection of dividing PBMC, presumably as it is during infection of MDMs.

Interaction and co-encapsidation of human immunodeficiency virus type 1 Gag and Vif recombinant proteins.

Human immunodeficiency virus type 1 (HIV-1) wild-type (WT) virion infectivity factor (Vif) protein (Vifwt) and full-length Gag precursor (Pr55Gag) were found to be co-encapsidated into extracellular, membrane-enveloped virus-like particles released by budding from Sf9 cells co-expressing the two recombinant proteins in trans, with an average copy number of 3.5+/-0.6 Vifwt per 100 Pr55Gag molecules. No preferential localization at the plasma membrane was observed for recombinant Vif in the absence of Gag expression, and a significant proportion of Vif accumulated within the nucleus. Two conserved motifs, W89RKRRY94 and P156KKIKP161, seemed to act as nuclear addressing signals. The Pr55Gag and Vifwt interacting domains were analysed by biopanning of a phage-displayed hexapeptide library. The Vif-binding domain, which spanned residues H421-T470 in Pr55Gag, corresponded to the C-terminal region of nucleocapsid (NC), including the second zinc finger, the intermediate spacer peptide sp2 and the N-terminal half of the p6 domain. Deletions in these Gag domains significantly decreased the Vif encapsidation efficiency, and complete deletion of NC abolished Vif encapsidation. In Vif, four discrete Gag-binding sites were identified, within residues T68-L81 (site I) and W89-P100 (site II) in the central domain, and within residues P162-R173 (III) and P177-M189 (IV) at the C terminus. Substitutions in site I and deletion of site IV were detrimental to Vif encapsidation, whereas substitution of basic residues for alanine in sites III and IV had a positive effect. The data suggest a direct intracellular Gag-Vif interaction and the occurrence of a Pr55Gag-mediated membrane-targeting pathway for Vif in Sf9 cells.

Visna virus dUTPase is dispensable for neuropathogenicity.

The major part of the dUTPase-encoding region of the visna virus genome was deleted. Intracerebral injection of the mutant virus resulted in a somewhat reduced viral load compared to that resulting from injection of the wild type, especially in the lungs, but the neuropathogenic effects were comparable. The dUTPase gene is dispensable for induction of lesions in the brain.

Human immunodeficiency virus type 1 Vif protein binds to the Pr55Gag precursor.

The Vif protein of human immunodeficiency virus type 1 is required for productive replication in peripheral blood lymphocytes. Previous reports suggest that vif-deleted viruses are limited in replication because of a defect in the late steps of the virus life cycle. One of the remaining questions is to determine whether the functional role of Vif involves a specific interaction with virus core proteins. In this study, we demonstrate a direct interaction between Vif and the Pr55Gag precursor in vitro as well as in infected cells. No interaction is observed between Vif and the mature capsid protein. The Pr55Gag-Vif interaction is detected (i) in the glutathione S-transferase system, with in vitro-translated proteins demonstrating a critical role of the NC p7 domain of the Gag precursor; (ii) with proteins expressed in infected cells; and (iii) by coimmunoprecipitation experiments. Deletion of the C-terminal 22 amino acids of Vif abolishes its interaction with the Pr55Gag precursor. Furthermore, point mutations in the C-terminal domain of Vif which have been previously shown to abolish virus infectivity and binding to cell membranes dramatically decrease the Gag-Vif interaction. These results suggest that the interaction between Vif and the pr55Gag precursor is a critical determinant of Vif function.

dUTPase-minus caprine arthritis-encephalitis virus is attenuated for pathogenesis and accumulates G-to-A substitutions.

The importance of the virally encoded dUTPase for CAEV replication, invasiveness, pathogenesis, and genetic stability was investigated in goats infected by viruses with single point (DU-G) and deletion (DU-1) mutations of the dUTPase gene (DU gene). The DU gene was found to be dispensable for CAEV replication in vivo as judged by times taken to seroconvert, frequencies of viral isolation, and tissue distribution of viral RNAs. DU- reversion at week 34 in one of three goats infected with the single point mutant DU-G, however, suggested that the viral dUTPase confers some advantages for replication in vivo. Moreover, we show that dUTPase is necessary for the timely development of bilateral arthritic lesions of the carpus. Finally, dUTPase was shown to efficiently prevent accumulation of G-to-A transitions in the viral genome.

Phenotypically Vif- human immunodeficiency virus type 1 is produced by chronically infected restrictive cells.

The permissivity of CD4+ transformed T cells for the replication of human immunodeficiency virus type 1 (HIV-1) vif mutants varies widely between different cell lines. Mutant vif-negative viruses propagate normally in permissive CD4+ cell lines but are unable to establish a productive infection in restrictive cell lines such as H9. As a consequence, elucidation of the function of Vif has been considerably hampered by the inherent difficulty in obtaining a stable source of authentically replication-defective vif-negative viral particles produced by restrictive cells. vif-negative, vpr-negative HIV-1 strain NDK stock, produced by the permissive SupT1 cell line, was used to infect restrictive H9 cells. By using a high multiplicity, infection of H9 cells was achieved, leading to persistent production of viral particles displaying a dramatically reduced infectious virus titer when measured in a single-cycle infectivity assay. Although these viral particles were unable to further propagate in H9 cells, they could replicate normally in CEM and SupT1 cells. Comparison of unprocessed and processed Gag proteins in the persistently produced vif-negative viral particles revealed no defect in the processing of polypeptide precursors, with no inversion of the Pr55gag/p24 ratio. In addition, there was no defect in Env incorporation for the vif-negative viral particles. Despite their apparently normal protein content, these particles were morphologically abnormal when examined by transmission electron microscopy, displaying a previously described abnormally condensed nucleoid. Chronically infected restrictive cell lines producing stable levels of phenotypically vif-negative HIV-1 particles could prove particularly useful in further studies on the function of Vif in the virus life cycle.

Requirement of caprine arthritis encephalitis virus vif gene for in vivo replication.

Replication of vif-caprine arthritis encephalitis virus (CAEV) is highly attenuated in primary goat synovial membrane cells and blood-derived macrophages compared to the wild-type (wt) virus. We investigated the requirement for CAEV Vif for in vivo replication and pathogenicity in goats by intra-articular injection of either infectious proviral DNA or viral supernatants. Wild-type CAEV DNA or virus inoculation induced persistent infection resulting in severe inflammatory arthritic lesions in the joints. We were unable to detect any sign of virus replication in vif- CAEV DNA inoculated goats, while vif- CAEV virus inoculation resulted in the seroconversion of the goats. However, virus isolation and RT-PCR analyses on blood-derived macrophage cultures remained negative throughout the experiment as well as in joint or lymphoid tissues taken at necropsy. No pathologic lesions could be observed in joint tissue sections examined at necropsy. Goats inoculated with the vif- virus demonstrated no protection against a pathogenic virus challenge. These results demonstrate that CAEV Vif is absolutely required for efficient in vivo virus replication and pathogenicity and provide additional evidence that live attenuated lentiviruses have to establish a persistent infection to induce efficient protective immunity.

The human immunodeficiency virus type 1 Vpr transactivator: cooperation with promoter-bound activator domains and binding to TFIIB.

Since the first report documenting that HIV-1 Vpr was involved in the stimulation of transactivation of several unrelated promoters, little additional information has been reported. By using transient transfection experiments, we confirmed and extended these previously reported data. Further in vivo experiments showed that Vpr can co-operatively stimulate transactivation activity of a minimal promoter containing one GAL4 DNA-binding site, when it is co-expressed with different heterologous activator domains fused to GAL4 DNA-binding domain. Thus, Vpr could transactivate in concert with an activator domain, but has no effect on the transactivation of a minimal promoter in the absence of activator protein. To investigate whether Vpr can interact with components of the basal transcriptional machinery, in vitro protein-protein binding assays were performed using either translated, radiolabeled Vpr or TFIIB proteins and glutathione S-transferase Vpr or TFIIB chimeric proteins. We demonstrated that the portion of Vpr ranging from amino acids 15 to 77 interacts specifically with the basal transcription factor TFIIB. Also, our data indicated that the N-terminal domain of TFIIB is required for the interaction.

Replication properties of dUTPase-deficient mutants of caprine and ovine lentiviruses.

The virion-associated dUTPase activities of caprine arthritis-encephalitis virus (CAEV) and visna virus were determined by using an assay which measure the actual ability of the dUTPase to prevent the dUTP misincorporations into cDNA during reverse transcription. We showed that the CAEV molecular clone from the Cork isolate was dUTPase defective as a result of a single amino acid substitution. Using this point mutant and deletion mutants of CAEV as well as a deletion mutant of visna virus, we demonstrated that dUTPase-deficient viruses replicate similarly to wild-type viruses in dividing cells but show delayed replication in nondividing primary macrophages.

Human immunodeficiency virus type 1 Vpr protein binds to the uracil DNA glycosylase DNA repair enzyme.

The role of the accessory gene product Vpr during human immunodeficiency virus type 1 infection remains unclear. We have used the yeast two-hybrid system to identify cellular proteins that interact with Vpr and could be involved in its function. A cDNA clone which encodes the human uracil DNA glycosylase (UNG), a DNA repair enzyme involved in removal of uracil in DNA, has been isolated. Interaction between Vpr and UNG has been demonstrated by in vitro protein-protein binding assays using translated, radiolabeled Vpr and UNG recombinant proteins expressed as a glutathione S-transferase fusion protein. Conversely, purified UNG has been demonstrated to interact with Vpr recombinant protein expressed as a glutathione S-transferase fusion protein. Coimmunoprecipitation experiments confirmed that Vpr and UNG are associated within cells expressing Vpr. By using a panel of C- and N-terminally deleted Vpr mutants, we have determined that the core protein of Vpr, spanning amino acids 15 to 77, is involved in the interaction with UNG. We also demonstrate by in vitro experiments that the enzymatic activity of UNG is retained upon interaction with Vpr.

The caprine arthritis encephalitis virus tat gene is dispensable for efficient viral replication in vitro and in vivo.

Caprine arthritis encephalitis virus (CAEV) is a lentivirus closely related to visna virus and more distantly to other lentiviruses, such as human immunodeficiency virus. The genomes of visna virus and CAEV contain a tat gene encoding a protein able to weakly transactivate its own long terminal repeat, suggesting that transactivation may be a dispensable function for viral replication. Three different tat gene mutants of an infectious molecular clone of CAEV were used to study their replication after transfection or infection of primary goat synovial membrane cells and of blood-derived mononuclear cells or macrophages. Our results showed no difference between replication of the wild type and either the complete tat deletion mutant or the tat stop point mutant, whereas slower growth kinetics and lower levels of expression of the partial tat deletion mutant that of the wild type were obtained in these cells. Quantitative PCR and reverse transcription-PCR analyses of the different steps of a single replicative cycle revealed an identical pattern of retrotranscription, transcription, and viral production, whereas time course analysis demonstrated that the intracellular level of viral genomic RNA was affected by the partial tat deletion at later time points. We then compared the infectious properties of the wild-type and tat mutant viruses in vivo by direct inoculation of proviral DNAs into the joints of goats. All the animals seroconverted between 27 and 70 days postinoculation. Moreover, we were able to isolate tat mutant CAEV from blood-derived macrophages that was still able to infect synovial membrane cells in vitro. This study clearly demonstrates that the tat gene of CAEV is dispensable for viral replication in vitro and in vivo.

The vif gene is essential for efficient replication of caprine arthritis encephalitis virus in goat synovial membrane cells and affects the late steps of the virus replication cycle.

Complex retrovirus genomes contain a variable number of accessory genes, among which is the vif gene. We investigated in vitro the role of the vif gene of caprine arthritis encephalitis virus (CAEV) by studying the phenotype of five vif mutants after infection of primary goat synovial membrane (GSM) cells and blood-derived monocytes/macrophages. Any deletion introduced into the vif gene resulted in slow and low viral replication and production of virions with an infectious titer lower than that of wild-type viral particles. The wild-type phenotype could be restored by the trans expression of the vif gene in a complementation assay. Quantitative PCR and reverse transcription-PCR analyses were performed in order to determine which stage of the replicative cycle was impaired by the vif deletion. Our results demonstrated that CAEV Vif did not act at the level of reverse transcription or transcription but rather at the late stage of virus formation and/or release, as lower amounts of virus were produced after a single replicative cycle. The vif-deleted CAEV produced after 24 h of infection was still able to infect GSM cells, indicating that the vif gene is not essential for virus infectivity but is required for efficient virus production.

Peripheral blood mononuclear cells produce normal amounts of defective Vif- human immunodeficiency virus type 1 particles which are restricted for the preretrotranscription steps.

Previous studies have demonstrated the absence of viral replication of Vif- mutants in stimulated primary blood mononuclear cells (PBMC). Human immunodeficiency virus type 1 strain NDK Vif- mutants were propagated on the semipermissive CEM cell line, and the viral stock obtained was compared with the wild-type virus during a single cycle in PBMC. The Vif- virus was able to enter PBMC with the same efficiency as the wild type, as demonstrated by quantification of the strong-stop cDNA, and retrotranscription was observed for both viruses within 4 h postinfection. Using a PCR assay with an Alu-long terminal repeat pair of primers, we detected integration for both the wild-type and Vif- viruses. We then used qualitative and quantitative reverse transcription-mediated PCR techniques to study the steady-state level of intracellular and extracellular viral RNAs. All mRNA species were detected in PBMC infected with the wild-type virus or with the Vif- virus 36 h postinfection. Furthermore, quantification of viral RNA released from infected cells demonstrated similar levels of virus produced after a unique cycle of replication. However, the Vif- virus obtained after one replication cycle in PBMC was unable to initiate retrotranscription in permissive target cells. These data strongly suggest that the failure to infect target cells is due to a defect in the formation of the viral particle in PBMC.