INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The process of retroviral integration, in which newly reverse-transcribed viral DNA is inserted into the host cell chromosome, is essential for a productive infection (13, 23, 32, 46, 48). Integration of human immunodeficiency virus (HIV) cDNA is mediated by a complex of both viral and cellular proteins closely associated with viral DNA that is known as the preintegration complex or PIC (2, 3, 5, 16, 30, 33, 38). HIV cDNA integration can be divided into three main steps: (i) 3'-end processing, involving the removal of a dinucleotide from the 3' termini of the linear viral DNA molecule; (ii) strand transfer, in which both 3' ends of the viral DNA are covalently linked to precleaved host cellular DNA; and (iii) gap repair, where the 5' ends of viral DNA are trimmed and then ligated to the host cell DNA following repair of gapped regions generated by the strand-transfer reaction (1, 11, 21, 42). Although gap repair is likely to be accomplished by cellular proteins (10), the 3'-end processing and strand-transfer reactions are primarily mediated by the viral integrase protein, IN (40). The catalytic core region of the integrase protein contains three spatially conserved, invariable amino acids (D₆₄, D₁₁₆, and E₁₅₂) that have been shown to be indispensable for activity and are thought to be key components of the catalytic site (12).

To date, high-throughput screening for potential integrase inhibitors has primarily been performed in cell-free systems using purified integrase either alone or within the context of a partially purified PIC (4, 17, 18, 24, 25, 29, 36). Since these assays can be designed to test for inhibition of either the formation of the initial stable complex, 3'-end processing, strand transfer, or disintegration (the reverse of strand transfer), they can both rapidly identify potential inhibitors and also provide preliminary evidence about their mode of action. However, inhibitors targeting the integrase protein and/or PICs identified in this manner are frequently cytotoxic or do not exhibit antiviral activities in cell culture (42).

Recently, a number of compounds identified in cell-free assays have been shown to inhibit viral replication in cell culture without displaying significant cytotoxicity (15, 26, 31, 39, 44, 45, 49, 50). AR177 (a G-quartet-containing oligonucleotide that forms highly stable intermolecular tetrad structures) and members of the bisaroyl hydrazine family of integrase inhibitors have been shown to inhibit in vitro integration reactions in the nanomolar and low micromolar ranges respectively (6, 37; N. Neamati et al., submitted for publication). Furthermore, AR177 was shown to inhibit syncytia formation and productive infection in cell culture, albeit at higher concentrations than those observed for integrase inhibition in cell-free assays (15, 39). In addition, a new class of integration inhibitors containing a diketo acid moiety has been described (14, 26). Acute infections performed in the presence of such compounds (L-731,988 and L-708,906) not only abolished productive infection but also resulted in the accumulation of large amounts of circular DNA forms incapable of integration. In addition, mutations conferring resistance to these drugs in cell culture consistently mapped to defined regions within the integrase protein. Although these results strongly suggested that the antiviral effect observed was due to a selective block of the integration process in infected cells, a direct evaluation of whether the drugs inhibited the accumulation of integrated HIV-1 DNA was not performed.

Using a modified nested Alu-PCR to quantify HIV provirus in cells (N. Vandegraaff, R. Kumar, C. J. Burrell, and P. Li, submitted for publication), we have established an assay that can be used to evaluate potential inhibitors identified in cell-free systems for their ability to inhibit the accumulation of integrated HIV-1 DNA following acute infection in cell culture. In this study, five compounds from four structurally diverse classes of inhibitors, which have all been reported to inhibit the HIV-1 integrase enzyme in cell-free assays, were tested for their ability to block integration of newly synthesized HIV-1 DNA into T-cell genomic DNA. The accumulation of extrachromosomal HIV DNA was also monitored to establish whether blocks to viral infection resulted from the specific inhibition of viral integration or inhibition of events at, or prior to, reverse transcription of the viral genome.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Cells and virus. The virus inoculum used for infection consisted of H3B cell culture medium that was clarified to remove cells and debris. The H3B cell line is a laboratory clone of H9 cells that are persistently infected with the human T-cell leukemia virus type IIIB (HIV_HXB2) strain of HIV (34). The virus titer of the inoculum was 3.16 × 10⁶ 50% tissue culture infective dose (TCID₅₀) ml. HuT-78 cells are a CD4⁺ T-lymphoblastoid cell line obtained from the National Institutes of Health (NIH) AIDS Research and Reference Reagent Program (22). ACH-2 and 8E5 clonal cell lines are T-cell lines persistently infected with HIV (8, 20) and were obtained from the NIH AIDS Research and Reference Reagent Program. All cells were maintained in RPMI 1640 medium supplemented with 10% fetal bovine serum, L-glutamine, penicillin (1.2 µg/ml), and gentamicin (1.6 µg/ml) at 37°C and 5% CO₂.

Drugs and cell cytotoxicity assays. The compounds 5,8-dihydroxynaphthoquinone and quercetin dihydrate were obtained from Aldrich. L-708,906 and lamivudine (3TC) were kind gifts from David Bourke, Department of Medicinal Chemistry, Victorian College of Pharmacy, Australia. L-731,988 and an additional sample of L-708,906 were obtained from the Department of Antiviral Research, Merck Research Laboratories, West Point, Pa., and L17 was synthesized in the Laboratory of Medicinal Chemistry, Division of Basic Sciences, National Cancer Institute, Bethesda, Md. AR177 was synthesized locally (Geneworks), and zidovudine (AZT) was obtained from Sigma. With the exception of AR177, all drugs were made up to 10 mM stocks in dimethyl sulfoxide and then diluted further in serum-free RPMI 1640 to the working concentration. AR177 was dissolved and diluted in phosphate-buffered saline. Working concentrations of all drugs used except quercetin dihydrate were based on concentrations shown to inhibit viral release following infection of T cells (9, 26, 31, 35, 39). Quercetin dihydrate was used at 50 µM, a concentration approximately fourfold higher than that shown to inhibit strand transfer in cell-free systems (12 µM).

Virus infection. HuT-78 cells were routinely subcultured at 5 × 10⁵/ml 16 h prior to infection to ensure cells were in the log phase of growth. AZT was preincubated with cells for 16 h prior to infection. All other drugs were preincubated with cells for 1 h prior to infection. Infection was initiated by incubation of cells with virus at a nominal multiplicity of infection (MOI) of 0.5 TCID₅₀ units per cell at 4°C for 30 min. Cells and virus were then spun at 2,500 × g for 1 h at 37°C after which cells were allowed to recover in prewarmed warmed fresh media containing relevant drugs for 15 min at 37°C. Under these conditions of centrifugal enhancement, the actual MOI has been reported as 10 times that of the nominal MOI (28, 34, 41). Infected cells were subsequently washed three times in media containing appropriate drugs to remove unbound virus and then plated in a 48-well tray at a density of 1 × 10⁶ cells/ml. Viral release was monitored over time by measuring the P24 concentrations in 1/50, 1/200, and 1/500 dilutions of the culture supernatant using a commercially available kit (NEN).

Preparation of integrated viral DNA copy number standards and DNA extraction procedures. The HA8 integrated proviral standards, chromosomal DNA samples, and extrachromosomal DNA samples were prepared as outlined elsewhere (Vandegraaff et al., submitted). Briefly, the HA8 copy number standard used is a mixture of equivalent amounts of chromosomal DNA extracted from known numbers of the H3B, ACH-2, and 8E5 persistently infected cell lines containing two, one, and one copies of integrated HIV DNA, respectively (8, 20, 34). HIRT pellet (chromosomal DNA) and HIRT supernatant (extrachromosomal DNA) extractions were essentially performed as originally published (27) in the presence of 0.5 mg of proteinase K (Merck) per ml. To minimize sodium dodecyl sulfate contamination of the DNA preparations, all ethanol precipitations were performed at room temperature. DNA preparations were resuspended in water at 5,000 cell-equivalents/µl and stored at 20°C until use.

PCR procedures. All PCRs were performed in a Perkin-Elmer GeneAmp PCR 9600 system. PCR amplification of the single-copy human -globin gene was used to estimate the DNA content of the chromosomal DNA preparations made. PCRs (25 µl) were performed using 50 cell-equivalents of chromosomal DNA in 1× PCR Buffer II (Perkin-Elmer), 2 mM MgCl₂, 0.2 mM concentrations of deoxynucleoside triphosphates (dNTPs) (Promega), 25 pmol of -glo 1 and 25 pmol of -glo 2 primers (Table 1) using 2.5 U of Amplitaq DNA polymerase. Reactions were cycled as follows: 94°C for 3 min; 25 cycles of 94°C for 45 s, 58°C for 30 s, 72°C for 45 s; and a final extension of 72°C for 10 min.

Analysis of PCR products. PCR products (10 µl) were subjected to electrophoresis on 8% polyacrylamide gels and then transferred (electroblot apparatus) onto Hybond N+ nylon filters (Amersham). After denaturation and fixation using 0.4 M NaOH, the filters were subjected to Southern hybridization using Ultrahyb (Ambion). The probes used to detect -globin, mitochondrial, nested Alu-, and GAG PCR products (Table 1) were labeled using [-³²P]dATP with a Megaprime kit (Amersham). Following Southern hybridization, bands were quantified using Phosphorimager ImageQuant analysis and a standard curve was generated from the PCR products arising from amplification of known amounts of the HA8 standards.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Seven compounds were examined for their effect on the accumulation of integrated HIV-1 DNA following acute infection of HuT-78 cells. With the exception of L17, the initial description of all drugs and preliminary cell-free data are available elsewhere (17-19, 26, 37, 39, 42, 43). L17 is a member of the bisaroyl hydrazine family of integrase inhibitors initially described by Zhao and coworkers (51) and consists of two sulfhydrylated aromatic ring structures spaced by an N-N linkage (Fig. 1). This compound was recently shown to inhibit integrase with a 50% inhibitory concentration (IC₅₀) of 20 µM in cell-free integration assays and the productive infection of T cells with an IC₅₀ of 5 µM (Neamati et al., submitted). All drugs except 5,8-dihydroxynaphthoquinone were noncytotoxic under infection conditions, even at concentrations fivefold higher than that used in the assay (see Table 2). The compound 5,8-dihydroxynaphthoquinone (IC₅₀ of 2.5 µM for strand transfer) was shown to be highly cytotoxic when used at concentrations above 1 µM and was therefore not subjected to further analysis. The in vitro activities against purified integrase, cytotoxicity, and concentrations of each drug used in this study are presented in Table 2.

View larger version (7K): [in this window] [in a new window]   FIG. 1.   Structure of L17.

Initially, duplicate infections were performed in the presence of either 10 µM L-708,906 or 10 µM L-731,988 (both containing a diketo acid moiety), or 50 µM quercetin dihydrate (a flavone). The inhibitors of reverse transcription, AZT and 3TC (used at concentrations of 10 µM), served as positive controls for inhibition of extrachromosomal HIV DNA synthesis prior to integration. In the absence of drug, infected cultures displayed extensive syncytia formation by 26 h post infection (p.i.) (data not shown) and high levels of supernatant P24 by 50 h p.i. (Fig. 2), indicating that a productive infection had occurred. In all samples, levels of P24 rose slightly from 2 h p.i. to 26 h p.i., possibly due to detachment of the virus inoculum from the surface of cells after binding during the infection procedure. With the exception of quercetin dihydrate, all drugs inhibited syncytia formation (data not shown) and P24 release into the culture supernatant at 50 h p.i. (Fig. 2).

View larger version (15K): [in this window] [in a new window]   FIG. 2.   Effect of five compounds on the levels of P24 released into culture supernatants at 2, 26, and 50 h following infection of HuT-78 cells with HIVHXB2.

To examine the accumulation of integrated HIV DNA in the presence of each drug, HIRT chromosomal preparations (27) were made from infected cells at 2, 26, and 50 h p.i. DNA was subjected to a modified nested Alu-PCR (7, 47) that specifically detects integrated HIV DNA forms. As expected, integrated HIV DNA was not detected in cultures treated with the reverse transcriptase inhibitors AZT and 3TC (Fig. 3, Integrated DNA, and Fig. 4A). Similarly, integrated DNA accumulation was not detected in the presence of either L-708,906 or L-731,988. Consistent with the P24 results, levels of integrated DNA observed in the presence of quercetin dihydrate at both 26 and 50 h p.i. were comparable to those observed for infections performed in the absence of drug. As a control, first-round PCR without the Alu164 primer was performed on the 50-h p.i., drug-free sample. The absence of a detectable signal confirmed that the signals observed at 50 h p.i. in the drug-free samples (Fig. 3) were derived from first-round PCR amplification of integrated HIV sequences and not the nested PCR amplification of any contaminating extrachromosomal forms present in the chromosomal DNA preparations (data not shown).

View larger version (24K): [in this window] [in a new window]   FIG. 3.   Effect of the potential integration inhibitors L-708, 906, L-731, 988, and quercetin dihydrate on levels of integrated and extrachromosomal HIV DNA following infection of HuT-78 cells with HIVHXB2. PCRs were performed on 1,000 cell-equivalents of HIRT pellet and supernatant fractions from duplicate cultures using the PCR protocols for the quantification of integrated and extrachromosomal DNA, respectively (see Materials and Methods). DNA levels in the presence of each potential integration inhibitor were compared with those detected in a control culture (No Drug) or after treatment with either AZT or 3TC, which block DNA synthesis prior to integration. Amplification of the single-copy -globin gene and mitochondrial DNA were used to control for the cell-equivalent amounts of chromosomal (HIRT pellet) and extrachromosomal (HIRT supernatant) DNA, respectively.

View larger version (24K): [in this window] [in a new window]   FIG. 4.   Graphical representation of data presented in Fig. 3. Graphed values are averages of duplicate samples. (A) Integrated DNA levels at 2, 26, and 50 h p.i.: values obtained by PhosphorImage analysis of Southern blots were adjusted based on -globin content. (B) Extrachromosomal DNA accumulation at 2, 26, and 50 h p.i., after adjustment for mitochondrial DNA content.

Since an absence of integrated HIV DNA might reflect either a specific inhibition of HIV DNA integration or a block in the HIV-1 replication cycle prior to integration, HIRT supernatant fractions (containing extrachromosomal DNA forms) from all samples were assayed using a GAG PCR protocol that detects extrachromosomal HIV DNA to establish whether reverse transcription was proceeding to completion. As expected, drug-free cultures and those infections performed in the presence of quercetin dihydrate exhibited significant amounts of reverse-transcribed products at 26 h p.i., whereas those in which infection was performed in the presence of AZT and 3TC displayed negligible levels (Extrachromosomal DNA in Fig. 3 and 4B). Both L-708,906 and L-731,988 also allowed the accumulation of extrachromosomal DNA by 26 h p.i., although at marginally lower amounts than that observed for drug-free cultures. Extrachromosomal DNA then increased substantially from 26 to 50 h p.i. in both drug-free cultures and cultures with quercetin dihydrate, while little further increase was seen in cultures containing L-708,906 and L-731,988 (Fig. 4B). Since infected cultures incubated in the absence of drug or the presence of quercetin dihydrate exhibited high levels of P24 by 50 h p.i. (Fig. 2) and extensive syncytia by 26 h p.i. (data not shown), the increases in extrachromosomal DNA observed after 26 h p.i. are likely to reflect de novo reverse transcription resulting from secondary infection of HuT-78 cells by progeny virus released from infected cells.

Both AR177 (an oligonucleotide inhibitor) and L17 (a salicylhydrazide) have been shown to inhibit HIV integrase in cell-free systems and to block productive HIV infection in cell culture (37, 39; Neamati et al., submitted). L17 and AR177, used at concentrations of 30 and 10 µM, respectively, inhibited both P24 release and syncytia formation even after 50 h p.i. (data not shown). Both of these drugs totally abolished the accumulation of integrated DNA forms (Fig. 5, Integrated DNA). However, they also inhibited the accumulation of extrachromosomal HIV DNA forms in infected cells (Fig. 5, Extrachromosomal DNA), indicating a block in the viral replication cycle at, or prior to, reverse transcription.

View larger version (20K): [in this window] [in a new window]   FIG. 5.   Effects of compounds L17 and AR177 on the levels of integrated and extrachromosomal HIV DNA following infection of HuT-78 cells with HIVHXB2. PCRs were performed on 1,000 cell-equivalents of DNA in triplicate from single cultures. DNA levels in the presence of each inhibitor were compared with levels obtained in a control culture (No Drug) or after treatment with AZT. DNA recovery was standardized by amplifying the single-copy -globin gene (HIRT pellets) or mitochondrial DNA (HIRT supernatants) as outlined for Fig. 3.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In this study, two diketo acid compounds (L-708,906 and L-731,988) inhibited the accumulation of integrated HIV-1 DNA without altering the synthesis of extrachromosomal HIV cDNA in the first round of viral replication. Although this result is consistent with inhibition of the viral integrase protein, the drugs could also be blocking transport of newly synthesized viral DNA to the nucleus. This possibility seemed unlikely since increased levels of circular viral DNA (used as a marker of viral entry into the nucleus) have been observed following infection in the presence of these drugs (26). Our results are in close agreement with previous reports indicating that viral reverse transcription is unaffected by these diketo compounds (26). It has also been shown that PICs isolated from cells infected in the presence of L-731,988 were unable to facilitate the integration of HIV DNA into a X174 DNA target substrate in a cell-free system (26). Taken together, these results indicate that L-708,906 and L-731,988 selectively block the HIV-1 integration reaction in cell culture.

Although shown in biochemical assays to inhibit the 3' processing and strand-transfer reactions at 20 and 12 µM, respectively (43), quercetin dihydrate (a weak DNA intercalator and topoisomerase 2 inhibitor) had no antiviral activity (at 50 µM) in our experiments, based on P24 release into the culture supernatant, syncytia formation, and the accumulation of both integrated and extrachromosomal viral DNA. This finding further confirms previous observations that compounds identified in cell-free assays do not necessarily inhibit integration in cell culture. Such compounds may be denied access or inefficiently transported to their primary site(s) of action within cells. Alternatively, interactions with unrelated components within the cell might degrade or sequester these compounds, making them unavailable to exert their effect.

Like AZT and 3TC, AR177 inhibited the accumulation of both integrated HIV DNA forms and extrachromosomal DNA, indicating a block in viral replication at, or prior to, reverse transcription. Our finding is consistent with recent studies showing that the primary target of AR177 is the viral gp120 protein (15) and underscores the importance of performing cell-based assays to define the precise targets of drugs within cells. AR177 has been shown to interfere with the binding of a monoclonal antibody raised against the V3 loop of gp120, and mutations that confer viral resistance to AR177 in cell culture map to residues within the loop regions of the gp120 protein (15). Along with our findings, these data suggest that the primary target of AR177 is the process of viral entry. However, it is worth noting that blocks in the viral replication cycle prior to integration and nuclear import could potentially result from an inhibition of viral entry, an inhibition of PIC assembly, or a direct effect on the viral reverse transcription process. Like AR177, L17 was shown to not only inhibit the accumulation of integrated HIV DNA but also that of reverse-transcribed product. Although this finding suggests that the primary viral target of this drug in cell culture is unlikely to be the process of integration, the precise target of L17 cannot be elucidated without further analysis. Furthermore, until mutations conferring viral resistance to this drug are mapped, the possibility that this drug inhibits viral replication both at, or prior to, reverse transcription as well as at integration cannot be eliminated.

In this study, we have described an efficient assay for monitoring the accumulation of integrated HIV DNA over time following infection of cells with HIV-1. When coupled with the quantitative detection of viral extrachromosomal DNA (both linear and circular forms), this assay can rapidly evaluate potential anti-integration drugs, identified in cell-free screening systems, for their ability to specifically block the HIV-1 integration process in cell culture. Similar experiments using peripheral blood mononuclear cells isolated from HIV-seronegative patients will provide further data on drug efficacy in cell culture. Furthermore, using a modification of this assay in which the cycle number of the nested PCR is increased, we have achieved a sensitivity of 10 copies of integrated HIV DNA per 2 × 10⁵ cells (data not shown). This is a sensitivity level sufficient to monitor the integrated viral load in patients.