Mechanisms Involved in Stimulation of Human Immunodeficiency Virus Type 1 Replication by Aminooxypentane RANTES

Abstract

Aminooxypentane (AOP)-RANTES is a potent inhibitor of nonsyncytium-inducing (NSI), CCR5-tropic (R5) human immunodeficiency virus type 1 (HIV-1) isolates. Although classical chemotactic responses are not induced in primary leukocytes by AOP-RANTES, recent studies suggest that a remnant of cell signaling occurs upon binding of receptor to this compound. We have detected a breakthrough of NSI/R5 replication from the inhibitory effects of high AOP-RANTES concentrations (<100 nM). A stimulation of different primary syncytium-inducing (SI), CXCR4-tropic (X4) HIV-1 isolates was also observed in the presence of AOP-RANTES. This stimulation was also observed after 110 h in PCR and RT-PCR for minus-strand strong-stop DNA and unspliced and multiply spliced RNA, respectively. However, there was significant variability between different SI/X4 or NSI/R5 HIV-1 isolates with regard to this AOP-RANTES-mediated stimulation or breakthrough, respectively. To further define the mechanism(s) responsible for this AOP-RANTES effect, we performed detailed retroviral replication studies with an NSI/R5 (B-92BR021) and SI/X4 (D-92UG021) HIV-1 isolate in the presence of the drug. Treatment of peripheral blood mononuclear cells with 125 nM AOP-RANTES and virus did not alter coreceptor expression, HIV-1 entry, reverse transcription, or mRNA transcription from the long terminal repeat, but it did result in increased HIV-1 integration. This AOP-RANTES-mediated increase in HIV-1 integration was diminished by treatment with pertussis toxin. Phosphorylation of the mitogen-activated protein kinase (MAPK) isoforms, extracellular signal-regulated kinase 1 (ERK1) and ERK2, was increased in a CD4⁺ CCR5⁺ U87 cell line treated with AOP-RANTES or with an NSI/R5 HIV-1 isolate. These findings suggest that AOP-RANTES may induce a MAPK/ERK signal transduction pathway upon binding to a G-protein-coupled receptor. MAPK/ERK1 and -2 appear to phosphorylate the HIV-1 preintegration complex, a step necessary for nuclear translocation and successful integration.

Entry of human immunodeficiency virus type 1 (HIV-1) into the host cell is mediated through the binding of HIV-1 gp120/gp41 envelope glycoproteins to the CD4 receptor and to a coreceptor (1, 9, 18, 22, 26, 30). Most primary HIV-1 isolates utilize CCR5 during asymptomatic disease and occasionally switch to CXCR4 receptor usage in late HIV-1 disease (15, 52), even though 14 seven transmembrane G-protein coupled receptors (e.g., CCR2b, CCR3, CCR7, and CCR8) can serve as HIV-1 coreceptors (7, 10, 21, 24, 25, 29, 34). Previous HIV-1 phenotypic designations of non-syncytium-inducing (NSI), macrophage-tropic or syncytium inducing (SI), T cell line-tropic generally correspond to and are now referred to as CCR5 (R5)- or CXCR4 (X4)-tropic, respectively (20).

The discovery that some chemokines (i.e., RANTES [regulated upon activation normal T-cell expressed and secreted], MIP-1α [macrophage inflammatory protein 1α], MIP-1β, and SDF-1α [stromal derived factor 1α]) can inhibit HIV-1 infection provided early evidence that chemokine receptors may mediate HIV-1 entry (13). Chemokines are a superfamily of small proteins involved in the inflammation response (62). A chemokine ligand-receptor interaction can induce chemotaxis, polarization, ion channel gating, and signal transduction pathways involved in the activation of specific leukocyte populations (35, 45, 60, 62). However, the findings that (i) a chemokine can bind to multiple G-protein coupled receptors (46, 62) and (ii) these receptors bind several chemokines as ligands reveal the considerable redundancy in this inflammation response (46, 62). Differential binding of MIP-1α, MIP-1β, and RANTES to CCR5 (46, 62) is highlighted by the finding that RANTES and MIP-1α are the most effective in blocking entry of NSI/R5 HIV-1 isolates (13). Thus, analogs of RANTES or MIP-1α appear to be the best candidates for anti-HIV-1 compounds (51). Unfortunately, inhibition of HIV-1 entry is not the only effect exerted by β-chemokines on HIV-1 replication (19, 28, 33, 39, 47, 56, 58). A signal transduced by chemokine ligand-receptor binding leads to an up-regulation of certain nuclear transcription factors and protein kinases, many of which can activate HIV-1 replication (2, 16, 35, 40, 45, 60). This may explain why β-chemokines (e.g., monocyte chemotactic protein 1 [MCP-1], MCP-3, and RANTES) and the α-chemokine SDF-1 can stimulate SI/X4 HIV-1 replication (19, 28, 33, 36, 39, 47, 56, 58).

A potent and effective chemokine analog would require dissociation of inhibitory activity from the potential HIV-1 stimulatory effects. One such analog, RANTES with an N-terminal serine residue replaced by the n-pentane of glyoxylic acid (AOP), is a potent inhibitor of NSI/R5 HIV-1 laboratory isolates but does not induce the chemotactic response characteristic of RANTES-receptor interactions (51). Subsequent studies did show Ca²⁺ influx in CCR5-positive cells treated with AOP-RANTES to levels similar to those induced by RANTES (45). Independent of AOP-RANTES-induced signaling, primary NSI/R5 HIV-1 isolates showed considerable variability in sensitivity to AOP-RANTES in peripheral blood mononuclear cells (PBMC) from a single HIV-negative donor (55). However, even a 30-fold difference in the AOP-RANTES concentrations required for 50% inhibition (IC₅₀) of different HIV-1 isolates (IC₅₀ ranges from 0.14 to 1.2 nM) could be overcome by treatment with micromolar concentrations of drug. As described in this and other studies (19, 28, 33, 39, 48), micromolar concentrations of AOP-RANTES would stimulate SI/X4 virus replication and permit a breakthrough of NSI/R5 HIV-1 isolates.

In this study, we have investigated the potential mechanisms for this AOP-RANTES-mediated breakthrough of NSI/R5 or stimulation of SI/X4 HIV-1 replication. NSI/R5 isolates showed variable sensitivities to both AOP-RANTES inhibition (55) and breakthrough, but there was no apparent relationship between the two effects. Variable stimulation of viral replication by AOP-RANTES analogs was also observed with different SI/X4 isolates, with other RANTES analogs, and in PBMC from different donors. Breakthrough of NSI/R5 or stimulation of SI/X4 by AOP-RANTES was not a consequence of increased CCR5 or CXCR4 expression, enhanced HIV-1 entry, increased reverse transcription, or activation of HIV-1 transcription by nuclear transcriptional factors. However, treatment with AOP-RANTES and RANTES did result in increased HIV-1 integration at 24 h, followed by a subsequent increase in proviral DNA and HIV-1 mRNA synthesis. Stimulation of SI/X4 HIV-1 replication by AOP-RANTES was diminished in PBMC pretreated with either pertussis toxin or PD98059 (MEK inhibitor). Furthermore, we did observe an increase in phosphorylated mitogen-activated protein kinase/extracellular signal-regulated kinase (MAPK/ERK) when U87/CD4/CCR5 cells were treated with AOP-RANTES or an NSI/R5 HIV-1 isolate. Findings from this study suggest that the MAPK/ERK signal transduction pathway is activated via binding of AOP-RANTES to a G-protein coupled receptor (e.g., CCR5). This pathway has been associated with an upregulation of several steps in HIV-1 replication (41, 49, 63, 64), including integration (31, 42).

MATERIALS AND METHODS

Cell culture.

PBMC from an HIV-1-seronegative donor were separated from heparinized blood by Ficoll-Paque density centrifugation and cultured in RPMI-1640 medium (Mediatech Inc.) supplemented with l-glutamine, 10% fetal bovine serum (Mediatech, Inc.), 10 mM HEPES buffer, 100-IU/ml and 100-μg/ml penicillin-streptomycin, 1 U of phytohemagglutinin/ml, and 1 ng of interleukin-2 (Gibco)/ml. The cells were suspended (2 × 10⁶ cells/ml) and grown for 3 days in culture at 37°C and 5% CO₂. U87, HeLa, and A549 cell lines obtained from the AIDS Research and Reference Reagent Program were grown in Dulbecco modified Eagle medium (DMEM) supplemented with 10% fetal calf serum, penicillin (100 U/ml), and streptomycin (100 μg/ml). U87/CD4 cells stably transfected with CCR1, CCR3, CCR5, and CXCR4 (5) were obtained from the American Type Culture Collection and grown in DMEM supplemented with 10% fetal calf serum, penicillin (100 U/ml), streptomycin (100 μg/ml), and G418 sulfate (1 mg/ml) at 37°C and 5% CO₂.

Viruses.

The NSI/R5 HIV-1 strains (A-92RW009, B-92BR021, B-92TH026, C-92BR025, E/A-92TH022, and B-Bal) and SI/X4 strains (B-HXB2, D-92UG021, A-92UG029, and E-CMU06) were obtained from the AIDS Research and Reagent Program for this study. The letter before the dash indicates the subtype of the viral envelope and is followed by the year of isolation, country of origin, and strain number (e.g., A-92RW009 is a clade A HIV-1 strain isolated in Rwanda in 1992). The viral stocks were prepared as previously described (43, 55). The 50% tissue culture infectivity dose values were calculated for each virus using the Reed-Muench technique (14).

Infection assays.

PBMC (10⁶ cells) were treated for 12 h with no drug, AOP-RANTES (312.5 nM or 2.5 μg/ml, 125 nM or 1 μg/ml, 31.25 nM or 0.25 μg/ml, and 0.313 nM or 2.5 ng/ml), or RANTES (6.25 nM or 50 ng/ml). Pertussis toxin (500 ng/ml) was also added 6 h prior to treatment with 125 nM AOP-RANTES. After 12 h, cultures were exposed to one of various NSI/R5 HIV-1 isolates (A-92RW009, B-92BR021, B-92TH026, C-92BR025, E-92TH022) or SI/X4 HIV-1 isolates (D-92UG021, A-92UG029, E-CMU06) at a multiplicity of infection (MOI) of 0.01. Samples of cells and cell culture supernatant were removed at 8, 24, and 110 h, 6 days, and 12 days postinfection. Virus production was monitored by reverse transcriptase (RT) activity in culture supernatants as described previously (55). To determine the effect of AOP-RANTES on entry, PBMC were placed into 24-well plates (10⁶ cells/well) and treated with AOP-RANTES (125, 31.25, and 0.313 nM) or RANTES (6.25 nM) 12 h prior to infection, 12 h postinfection, or at the time of infection with an NSI/R5 HIV-1 isolate (B-92BR021). Cells and culture supernatant were removed 24 h, 110 h, 6 days, and 10 days postinfection.

PCR and reverse transcription reaction.

PBMC were lysed with a solution containing 0.1 mg of gelatin/ml, 50 mM NaCl, 10 mM Tris (pH 8.3), 2.5 mM MgCl₂, 0.45% NP-40, and 0.45% Tween 20, treated with 1 mg of proteinase K (Gibco)/ml, and incubated for 1 h at 60°C. Cellular and viral DNA was extracted with phenol-chloroform and precipitated in 70% ethanol as described previously (43). HIV-1 minus-strand strong-stop DNA was PCR amplified from these samples using the A13 and ³²P-end-labeled S1 primer pairs as described previously (4). Tenfold dilutions of B-HXB2 DNA (10 to 10⁶ copies) were also PCR amplified as an amplification control. For an internal control, a region of mitochondrial DNA was amplified using the MTA and MTS primer pair as described previously (4). RNA was extracted from lysed PBMC using the RNeasy minikit (QIAGEN Inc.) and then reverse transcribed using random hexamer primers (Gibco) and Superscript II murine leukemia virus RT (Roche). The cDNA was then PCR amplified using primers specific for unspliced (SUNS-7 and AUNS-7) and multiply spliced (SMS-7 and AMS-7) HIV-1 RNA as previously described (32). To control for concentration of input RNA and reverse transcription, the SBA-7 and ABA-7 primer pair was used to PCR amplify β-globin cDNA (32).

Fluorescence-activated cell sorter (FACS) analysis.

Unstimulated PBMC were treated for 12 h with RANTES (6.25 nM), AOP-RANTES (125 or 0.313 nM), or AOP-RANTES (125 nM) with 500 ng of pertussis toxin (Gibco)/ml. Pertussis toxin was added 6 h prior to the addition of 125 nM AOP-RANTES. After treatment, cells were pelleted by centrifugation at 800 × g for 10 min, resuspended in phosphate-buffered saline (PBS) containing 5% bovine serum albumin, and incubated on ice for 15 min. Cells were again pelleted and resuspended in 50 μl of PBS prior to the addition of antibodies. Five microliters of peridinin chlorophyll protein-conjugated anti-human CD4 antibody plus 5 μl of phycoerythrin (PE)-conjugated anti-human CXCR4 antibody, 20 μl of PE-conjugated anti-human CCR5 antibody, and 5 μl of PE-conjugated mouse immunoglobulin G2a (IgG2a), κ isotype standard) (PharmMingen), was added to the cells followed by incubation in the dark for 30 min on ice. Cells were then washed with 5% bovine serum albumin–PBS and 500 μl of PBS. After the final wash, cells were fixed with 300 μl of 1% paraformaldehyde and analyzed using a FacScan flow cytometer and Lysis II software (Becton Dickinson).

Plasmid construction.

The long terminal repeat (LTR) region of the HIV-1 D-92UG021 strain was PCR amplified from DNA of infected PBMC using LTR1 and LTR2 as external primers and S2-LTR4 and LTR3-LTR4 as nested primer pairs (44). DNA was ethanol precipitated, resuspended in 40 μl of H₂O, and then ligated into the pCR-TOPO vector using the TOPO TA Cloning kit (Invitrogen). Plasmid DNA containing the D-92UG021 LTR was purified using the Qiagen Plasmid Mini Kit (Qiagen Inc.). An Asp718-to-XhoI (Boehringer Mannheim) fragment (532 bp) containing the U3-R region of the LTR was cut, purified as described above, and then subcloned upstream of the luciferase gene in the pGL3 Basic Vector (Promega Corp.). Plasmids containing the HXB2 and D-92UG021 LTRs were designated pLTR_HXB2Luc and pLTR_D-92UG021Luc, respectively.

Cell transfection and luciferase assay.

Twenty-four hours prior to transfection, A549, HeLa, U87/CD4/CCR5, and U87/CD4/CXCR4 cells were split into media containing no antibiotics. Cells (18 × 10⁶) were transfected with 24 μg of the pLTR_D-92UG021Luc or pLTR_HXB2Luc vectors using Fugene 6 transfection reagent (Roche). Transfected cells were treated with AOP-RANTES (125 nM, 0.313 nM) or RANTES (6.25 nM) for 12 h. In one sample, cells were pretreated with 500 ng of pertussis toxin/ml for 6 h and then exposed to 125 nM AOP-RANTES. Tumor necrosis factor alpha (TNF-α) (100 ng/ml) was used as a positive control for LTR activation. Luciferase was extracted from cells using the Reporter Lysis Buffer (Promega Corp.) and stored at −70°C. Extract (20 μl) was added to Luciferase Assay Reagent (100 μl) (Promega Corp.) and read in the Monolight 2010 luminometer (Analytical Luminescence Laboratory).

Immunoprecipitations and Western blot analyses.

U87/CD4/CCR5 cells (5 × 10⁶ cells per condition) were resuspended in serum-free DMEM and left untreated or treated with 500 ng of pertussis toxin/ml or with 50 μM PD98059, an inhibitor of the MAPK/ERK pathway, for 6 h prior to the addition of 125 nM AOP-RANTES. Tetradecanoylphorbol 13-acetate–phorbol 12-myristate 13-acetate (TPA/PMA) (50 ng/ml) was used as a positive control. Cells were harvested at 5 min and at 2, 5, and 24 h. Phosphorylated ERK1 and -2 isoforms were immunoprecipitated from the cell lysates with the p-ERK antibody (Santa Cruz Biotechnology Inc.) using an immunoprecipitation kit (Protein A) (Roche). Immunoprecipitations were performed according to the manufacturer's protocol. The immunoprecipitated proteins were separated by gel electrophoresis on a sodium dodecyl sulfate–10% polyacrylamide gel and then transferred onto a nitrocellulose membrane. Blots were probed with the antibody specific for the phosphorylated forms of ERK1 and -2 (p-ERK) (Santa Cruz Biotechnology Inc.), developed by electrochemiluminescence using SuperSignal West Pico Chemiluminescent Substrate (Pierce), and exposed to film.

Integration assay.

This PCR-based assay was designed to detect HIV-1 DNA integrated upstream of an Alu sequence in the host cell genome (11). DNA samples from the infection assays were added to the external PCRs along with the Alu and Alu-LTR primers. The Alu primer anneals to the highly redundant Alu sequence in the genome, while the Alu-LTR anneals to sequence in the U3 region. The PCR cycling conditions (35 cycles with a 72°C polymerization step for 2 min) were optimized for amplification of a 2-kbp fragment (i.e., HIV-1 DNA integrated within 2 kbp of an Alu sequence in the host cell genome). A nested PCR was then performed on the integrated HIV-1 DNA amplified in the external PCR. Three microliters of external amplification, cold S1 primer, and γ-³²P-end-labeled A2 primer were added to the nested reaction mixture. To ensure that unintegrated DNA was not amplified in this PCR, 0.18 μl (or the equivalent amount of the original DNA sample carried over from the external to the nested amplification) was PCR amplified with the nested primer pair (S1-γ-³²P-labeled A2). Tenfold dilutions of B-HXB2 DNA were also amplified with the nested primers as an amplification control.

RESULTS

AOP-RANTES has a dichotomous effect on HIV-1 replication in PBMC.

To examine the effects of AOP-RANTES on replication of NSI/R5 HIV-1 isolates, the compound was added to PBMC about 4 h prior to the addition of HIV-1 R5 isolates (B-Bal, A-92RW009, B-92BR021, B-92TH026, C-92BR025, and E-92TH022). Virus production in the culture supernatant was measured 6 and 10 days postinfection using a radioactive RT assay (Fig. 1). The concentration of AOP-RANTES required for IC₅₀ was previously determined with PBMC treated simultaneously with virus and drug (55). Although simultaneous treatment significantly delayed any breakthrough of NSI/R5 replication (see below), inhibition by AOP-RANTES did not differ in conditions where the drug was added prior to or during the addition of virus. High concentrations of AOP-RANTES (>125 nM) were necessary for primary NSI/R5 HIV-1 to break through the inhibitory effects of AOP-RANTES (Fig. 1A). At a 313 nM concentration of AOP-RANTES, this breakthrough was significant considering that the replication of the A-92RW009 and E-92TH022 isolates approached levels observed in the absence of drug. However, breakthrough in the presence of high AOP-RANTES concentrations was quite variable among primary NSI/R5 isolates. There was no apparent correlation between breakthrough and the sensitivity of these isolates to AOP-RANTES inhibition (Fig. 1A) (55). This lack of correlation suggests that independent mechanisms are responsible for AOP-RANTES inhibition and stimulation of HIV-1 replication in PBMC cultures. In the case of RANTES, it was difficult to discern a difference between inhibition (10- to 100-fold less potent than AOP-RANTES) and possible stimulation of NSI/R5 replication.

FIG. 1.

AOP-RANTES and RANTES induce breakthrough of NSI/R5 HIV-1 replication and activation of SI/X4 HIV-1 replication. PBMC were pretreated with no drug, AOP-RANTES (0.031, 0.31, 3.1, 31.3, 125, or 313 nM), or RANTES (6.25 nM) for 4 h and then exposed to various HIV-1 strains listed in the figure. (A) Inhibition of NSI/R5 HIV-1 isolates at lower AOP-RANTES concentrations and then a breakthrough at higher drug concentrations. Virus production was measured by RT activity in the cell culture supernatants at day 10 postinfection. All RT activities were then plotted relative to the infections in the absence of drug. (B) Effects of AOP-RANTES and RANTES on replication of SI/X4 isolates. PBMC were treated with AOP-RANTES (0.31 or 125 nM) or RANTES (6.3 nM) for 12 h or pretreated with pertussis toxin (P.Tx.) prior to addition of 125 nM AOP-RANTES and virus. (C) Plot of fold replication, relative to that of the control, of the SI/X4 isolate, D-92UG021, in the presence of 125 nM AOP-RANTES and in PBMC of three independent donors.

The dichotomous effects of AOP-RANTES on NSI/R5 isolates confound any detailed study of the possible mechanisms involved in AOP-RANTES stimulation of HIV-1 replication. Thus, use of HIV-1 X4 isolates (B-HXB2, D-92UG021, A-92UG029, and E-CMU06) was more practical for investigation of the stimulation effects mediated by AOP-RANTES (Fig. 1B). AOP-RANTES does not bind the CXCR4 receptor but could have indirect effects on HIV-1 replication via interactions with CCR5, CCR3, CCR1, or even membrane proteins with glycosylaminoglycan moieties (61, 62). Previous studies suggest that AOP-RANTES and RANTES may induce the strongest signals through the CCR5 receptor (51, 62).

AOP-RANTES was added to PBMC 12 h prior to the addition of HIV-1 SI/X4 isolates, in contrast to the 4-h preincubation for Fig. 1A. Viral production in culture supernatant was quantified by a radioactive RT assay on days 7 and 10 postinfection. Similar to the AOP-RANTES-mediated breakthrough of NSI/R5 replication, treatment with AOP-RANTES resulted in variable stimulation of four different SI/X4 primary HIV-1 isolates. Although the amount of virus did increase between days 7 and 10, the fold increase in virus production in the presence of AOP-RANTES or RANTES, in proportion to no-drug conditions, did not change during this time. The SI/X4 isolate, D-92UG021, responded with the greatest increase in replication (5.2 times greater than that for the untreated control). Interestingly, laboratory isolates, B-BaL (NSI/R5) (Fig. 1A) and B-HXB2 (SI/X4) (Fig. 1B), were the least responsive to AOP-RANTES-mediated breakthrough or stimulation, respectively. Stimulation by AOP-RANTES was diminished by pretreatment of PBMC with 500 ng of pertussis toxin/ml, suggesting transduction of a signal through a G-protein coupled receptor (Fig. 1B). As reported by others (19, 28, 33, 39), RANTES appears to induce stimulation of SI/X4 HIV-1 replication at a concentration considerably lower (6.3 nM) than that required for the AOP-RANTES effect (Fig. 1B). Stimulation by RANTES (6.3 nM) was also pertussis toxin sensitive (data not shown). Variable stimulation by 125 nM AOP-RANTES was observed with different primary HIV-1 isolates (1.4-fold with A-92UG029 to 5.2-fold with D-92UG021). The level of AOP-RANTES stimulation could be increased by using less virus (i.e., low MOIs of 0.001) for the initial infection and by increasing the incubation time of cells with the drug. We and others (28) have optimized for incubation time and determined that peak stimulation was observed after 12 to 24 h. However, all subsequent experiments required a higher MOI (0.01) for accurate detection and quantitation of virus-specific products during replication. Amounts of these virus-specific products at 8, 24, and 110 h were then compared to the virus release after 10 days in the presence of AOP-RANTES (see below). As described in the legends to Fig. 1A and B, differences in AOP-RANTES-mediated stimulation or breakthrough were evident between different SI/X4 or NSI/R5 primary isolates. It is important to note that PBMC from the same donor and blood draw were used for all experiments. In the next set of infections, we determined whether host variations could affect HIV-1 stimulation by AOP-RANTES. PBMC from three different donors were pretreated with AOP-RANTES (125 nM) and then exposed to the D-92UG021 SI/X4 isolate. Although AOP-RANTES (125 nM) did increase D-92UG021 replication in all PBMC, the level of stimulation varied from a 1.8-fold to a 5.2-fold increase over that for the untreated infections (Fig. 1C). Variations in AOP-RANTES stimulation did not appear to correspond with CCR5 expression on PBMC (data not shown). However, this limited data set does not adequately address the impact of these or other host factors on host variation to AOP-RANTES stimulation. It is important to note that this host variation is distinct from the variable stimulation by AOP-RANTES observed among different primary HIV-1 isolates in the same PBMC cultures (Fig. 1A and B).

Pretreatment with AOP-RANTES prior to virus infection may be necessary for breakthrough or stimulation.

A short preincubation (4 h) of PBMC with high concentrations of AOP-RANTES (<100 nM) resulted in an eventual breakthrough of NSI/R5 HIV-1 replication (Fig. 1A). Addition of AOP-RANTES to PBMC 12 h prior to the addition of an SI/X4 isolate (D-92UG021) resulted in a significant stimulation of replication. To determine if the time of AOP-RANTES treatment affects breakthrough of NSI/R5 HIV-1, drug was added to PBMC 12 h before, 12 h after, or simultaneously with an NSI/R5 HIV-1 isolate (B-92BR021). HIV-1 minus-strand strong-stop DNA was PCR amplified from DNA extracts of 24-h (data not shown) and 110-h samples (Fig. 2A). A 0.01 MOI of the NSI/R5 isolate was insufficient for PCR detection of minus-strand strong-stop DNA at 24 h. At 110 h, inhibition of minus-strand strong-stop DNA was apparent in all HIV-1 infections treated with AOP-RANTES or RANTES. Increasing concentrations of AOP-RANTES resulted in a decrease of minus-strand strong-stop DNA, but no change was found in the amount of mitochondrial DNA detected by PCR. Preincubation with AOP-RANTES resulted in a greater inhibition of minus-strand strong-stop DNA synthesis than did simultaneous addition of virus and drug or treatment of virus and then drug (Fig. 2A). A breakthrough of minus-strand strong-stop DNA synthesis, as opposed to inhibition by AOP-RANTES, was evident in the sample treated with a high concentration (125 nM) of AOP-RANTES prior to virus exposure (Fig. 2A, lane 7). The same concentration of AOP-RANTES, added simultaneously with virus or 12 h after virus, resulted in complete inhibition of minus-strand strong-stop DNA synthesis at 110 h (Fig. 2A, lanes 3 and 11).

FIG. 2.

Quantifying minus-strand strong-stop DNA synthesis and virus production in PBMC treated with AOP-RANTES and exposed to the NSI/R5 HIV-1 B-92BR021. This experiment involved three conditions: (i) the addition of HIV-1 B-92BR021 12 h prior to treatment with AOP-RANTES (0.31, 3.1, or 125 nM) or RANTES (6.3 nM), (ii) the addition of drug and then virus, and (iii) simultaneous exposure to virus and drug. Mitochondrial DNA and minus-strand strong-stop DNA were amplified from DNA extracts of the 24- and 110-h cellular lysates. Minus-strand strong-stop DNA was detected and PCR amplified only in the 110-h samples (A). Tenfold dilutions (10⁶ to 10 copies) of B-HXB2 DNA were employed as a DNA template for the PCR amplification of minus-strand strong-stop DNA and as a positive control. (B) Plot showing the relative production of the primary NSI/R5 HIV-1 isolate, B-92BR021, at day 10 postinfection under the conditions described above. Virus production was measured in culture supernatants by an RT assay.

In general, virus production at day 10 correlated with the amount of minus-strand strong-stop DNA that was PCR amplified postinfection (Fig. 2B). For example, treatment with 31.3 or 0.31 nM AOP-RANTES resulted in both inhibition of minus-strand strong-stop DNA synthesis and an equivalent decrease in virus production, compared to results with the untreated controls. At day 10 postinfection, breakthrough of HIV-1 B-92BR021 replication was again evident in PBMC cultures pretreated with 125 nM AOP-RANTES (Fig. 2B). In the sample treated simultaneously with 125 nM AOP-RANTES and virus, there was a discrepancy between the lack of minus-strand strong-stop DNA in the 110-h sample (Fig. 2A, lane 11) and a breakthrough in virus production at day 10 (Fig. 2B). This is probably due to a delay in the signal responsible for stimulation by AOP-RANTES. The level of B-92BR021 breakthrough when 125 nM AOP-RANTES was added simultaneously with virus was similar to that breakthrough observed when PBMC were preincubated for 4 h with drug (see Fig. 1A). Addition of AOP-RANTES after HIV-1 infection displayed no stimulation of viral replication.

AOP-RANTES stimulation of viral replication appears to be independent of entry and occurs after reverse transcription.

An SI/X4 HIV-1 isolate was used to investigate the possibility that AOP-RANTES could induce cell signaling linked to a subsequent increase in HIV-1 replication. PBMC were treated for 12 h with AOP-RANTES or RANTES prior to infection with D-92UG021 (the SI isolate having the greatest stimulation by AOP-RANTES) (Fig. 1B). PCR amplification for minus-strand strong-stop DNA and RT-PCR amplification for spliced and unspliced HIV-1 RNA were performed on the 8-, 24-, and 110-h samples. De novo synthesis of minus-strand strong-stop DNA was detected at 8 h postinfection by PCR (Fig. 3A) and was at least 100-fold greater than the low levels of minus-strand strong-stop DNA found in cell-free virus (3). However, there was no difference between HIV-1 infections pretreated with or without AOP-RANTES, suggesting that entry was not responsible for the subsequent HIV-1 activation by AOP-RANTES (Fig. 3A). The amounts of minus-strand strong-stop DNA in the 8- and 24-h samples were nearly identical (data not shown). After 110 h, an increase in the amount of PCR-amplified minus-strand strong-stop DNA was apparent in samples treated with AOP-RANTES or RANTES compared to results with the untreated control (Fig. 3B). Treatment of PBMC with 500 ng of pertussis toxin/ml and then 125 nM AOP-RANTES, compared to 125 nM AOP-RANTES alone, did reduce the amount of minus-strand strong-stop DNA at 110 h postinfection, suggesting the possible involvement of a G-protein coupled receptor. The effect of each AOP-RANTES and RANTES treatment on minus-strand strong-stop DNA synthesis at 110 h was similar to the stimulation of virus production at day 10 postinfection, as shown in Fig. 3C.

FIG. 3.

Amplification of minus-strand strong-stop DNA from PBMC treated with AOP-RANTES and the SI/X4 isolate D-92UG021. PBMC were treated with AOP-RANTES (0.31 or 125 nM) or RANTES (6.3 nM) for 12 h prior to the addition of the SI/X4 D-92UG021 isolate or treated with virus and then drug. In one condition, cells were treated with pertussis toxin (P.Tx.) (500 ng/ml) for 6 h prior to the addition of AOP-RANTES. Infections were harvested at various time points (8, 24, and 110 h), lysed, and subjected to PCR amplifications with radiolabeled primers specific for minus-strand strong-stop or mitochondrial DNA. The amplification control for minus-strand strong-stop DNA is shown in Fig. 2. Panels A and B show the results from the 8-h and 110-h time points, respectively. Similar amounts of minus-strand strong-stop DNA were observed in the 8- and 24-h samples. neg, negative. (C) Comparison of the amplification of minus-strand strong-stop DNA at 110 h to virus production 10 days postinfection (see the data in Fig. 1B), both of which are relative to the no-drug control.

To examine the effects of AOP-RANTES and RANTES on HIV-1 RNA synthesis, RNA was extracted from the 24- and 110-h samples and subjected to reverse transcription and PCR amplification using primers specific for unspliced and multiply spliced HIV-1 mRNA transcripts. After 24 h, a relatively equal amount of unspliced HIV-1 RNA was detected by RT-PCR amplification in samples treated with or without AOP-RANTES. Little or no multiply spliced mRNA was RT-PCR amplified at 24 h (Fig. 4A). At 110 h postinfection, increased production of multiply spliced HIV-1 RNA was evident by RT-PCR in samples treated with AOP-RANTES and RANTES (Fig. 4B). As indicated by the β-globin mRNA amplification, there was a twofold decrease in RNA added to the RT-PCR from the no-drug samples compared to the treated sample. However, the difference in HIV-1 multiply spliced mRNA amplification between these two samples was at least 10-fold. This increase was comparable to that of minus-strand strong-stop DNA in the presence of AOP-RANTES at 110 h and to the increase in D-92UG021 HIV-1 replication (Fig. 4C). Pretreatment of PBMC with pertussis toxin diminished the increase of multiply spliced HIV-1 mRNA mediated by 125 nM AOP-RANTES. All of the changes in multiply spliced mRNA expression at 110 h under each treatment condition reflected the effects on HIV-1 replication 10 days postinfection (Fig. 4C). Treatment with AOP-RANTES and RANTES resulted in little or no increase in unspliced HIV-1 mRNA (Fig. 4).

FIG. 4.

RT-PCR amplification of unspliced and multiply spliced HIV-1 mRNA from PBMC treated with AOP-RANTES and the SI/X4 isolate D-92UG021. PBMC were treated with AOP-RANTES (0.31 or 125 nM) or RANTES (6.3 nM) for 12 h prior to the addition of the SI/X4 D-92UG021 isolate. In one condition, cells were treated with pertussis toxin (P.Tx.) as described in the legend to Fig. 3. Infected cells were harvested at 24 and 110 h, lysed, and subjected to RT-PCR specific for unspliced HIV-1 RNA, multiply spliced HIV-1 mRNA, and β-globin mRNA. Panels A and B show the results from the 24-h and 110-h time points, respectively. (C) Comparison of the RT-PCR amplification of unspliced and multispliced HIV-1 mRNA to virus production 10 days postinfection (see data in Fig. 1B), both of which are relative to the no-drug control.

Effects of AOP-RANTES and RANTES on chemokine receptor expression.

Although results shown in Fig. 3 provide no evidence for increased entry of HIV-1 in the presence of AOP-RANTES, an earlier study suggests that treatment of cells with β-chemokines may increase CXCR4 mRNA expression (19). We examined the effects of AOP-RANTES or RANTES treatment on CCR5 and CXCR4 surface expression in PBMC (Fig. 5). PBMC were treated for 12 h with 0.31 nM AOP-RANTES, 125 nM AOP-RANTES with or without 500 ng of pertussis toxin/ml, or 6.25 nM RANTES. FACS analysis of cells stained with PE-conjugated anti-human CCR5 antibody indicate that a 10-fold increase in the AOP-RANTES concentration (Fig. 5A, panels II to IV) resulted in a 10-fold decrease in CCR5 surface expression or in the percentage of cells expressing CCR5 on the cell surface. This decrease in surface expression was unaffected by the addition of pertussis toxin (Fig. 5A, panel V). Downregulation of CCR5 also occurred in the presence of RANTES (Fig. 5A, panel VI). Although receptor downregulation was evident, RNase protection assays (RPA) specific for CCR5 message (Riboquant kit from PharMingen) showed no change in CCR5 mRNA expression during 48 h of incubation with AOP-RANTES (0.31 or 125 nM) or RANTES (6.3 nM). RPA were performed on 30-min and 3- and 48-h samples (data not shown). As indicated in Fig. 5A, a slight decrease in CCR5 surface expression in the presence of 0.31 nM AOP-RANTES or 6.3 nM RANTES did not necessarily correspond to inhibition of B-92BR021 HIV-1 (Fig. 5C) or the other NSI/R5 HIV-1 isolates (Fig. 1A). There is now evidence that competitive binding of CCR5 between virus and chemokine may be responsible for inhibition, rather than simply receptor downregulation upon binding to a chemokine (55). Higher concentrations of AOP-RANTES (125 nM) resulted in a further decrease in CCR5 surface detection and a breakthrough of NSI/R5 HIV-1 replication (Fig. 5C). Although this response appears counterintuitive, recent studies suggest that AOP-RANTES-CCR5 complexes are rapidly internalized, recycled to the cell surface, and then immediately reinternalized (50). This may permit continual restimulation of an AOP-RANTES-CCR5 signal transduction pathway that may activate HIV-1 replication (see below). In addition, the NSI/R5 isolate could have continual access to the reappearing CCR5 receptor on the cell surface.

FIG. 5.

Surface expression of CCR5 and CXCR4 in PBMC treated with AOP-RANTES and RANTES. PBMC were exposed to AOP-RANTES or RANTES for 12 h and then costained with PerCP-labeled anti-CD4 antibody and PE-labeled anti-CCR5 or anti-CXCR4 antibody. Untreated PBMC were also labeled with the anti-CD4 antibody and a conjugated mouse IgG2a κ isotype standard. Labeled cells were then analyzed using the FacScan flow cytometer and Lysis II software. Panels I to VI represent CCR5 (A) or CXCR4 (B) events in the CD4 positive-gated lymphocyte population. (C and D) Comparison of CCR5 or CXCR4 surface expression to NSI/R5 B-92BR021 and SI/X4 D-92UG021 virus production (day 10), respectively. P.Tx., pertussis toxin.

In contrast to CCR5, the percentage of PBMC expressing the CXCR4 receptor increased slightly, from 45.3% in the absence of drug (Fig. 5B, panel II) to 60.7% in the presence of 125 nM AOP-RANTES (Fig. 5B, panel IV). It is unlikely that a 15% increase in cells expressing the CXCR4 receptor could account for a fivefold increase of viral replication (Fig. 5D). In addition, no differences have been observed in viral entry (Fig. 3A). Although the presence of AOP-RANTES resulted in a slight increase in CXCR4 surface expression, RPA did not reveal a significant change in CXCR4 mRNA expression during the 48-h incubation with AOP-RANTES or RANTES (data not shown).

AOP-RANTES induces phosphorylation of MAPK/ERK.

Previous studies have shown an activation of the MAPK/ERK pathway upon CXCR4 binding to an SI/X4 HIV-1 isolate, soluble gp120, or SDF-1α (40, 42). A recent study has suggested that phosphorylation of MAPK/ERK may also be activated through the CCR5 receptor and could lead to an upregulation of HIV-1 replication (40). To investigate a possible effect on the MAPK/ERK pathway, PBMC were treated with AOP-RANTES, along with pertussis toxin (500 ng/ml) or PD98059 (50 μM), and then exposed to the HIV-1 isolate D-92UG021 (Fig. 6A). PD98059 is a weak inhibitor of MAPK/ERK phosphorylation by MEK1 and -2 kinase (23). This did translate to a weak PD98059 inhibition of AOP-RANTES-mediated stimulation of HIV-1 replication. HIV-1 stimulation by AOP-RANTES was effectively blocked by the addition of pertussis toxin but was minimally inhibited by PD98059 (Fig. 6A). This experiment provided weak but consistent evidence for the role of the MEK/ERK signaling cascade in AOP-RANTES-mediated stimulation of HIV-1. Based on these experiments, phosphorylation of MAPK/ERK was initially assessed in PBMC treated with AOP-RANTES by immunoprecipitation with an antibody specific for the phosphorylated p44 ERK1 and p42 ERK2 isoforms. However, AOP-RANTES did not induce significant levels of phosphorylated MAPK/ERK in PBMC (data not shown). This may be due to the weak signal transduced by AOP-RANTES and the relatively low percentage of cells expressing RANTES receptors in the PBMC cultures. However, a significant increase in phosphorylated MAPK/ERK was observed in U87 cells expressing CCR5 and CD4 when they were treated with 125 nM AOP-RANTES (Fig. 6B). This increase was observed 5 min after AOP-RANTES treatment (125 nM) and was sustained for at least 24 h (Fig. 6B). TPA/PMA (50 ng/ml) treatment of U87/CD4/CCR5 cells resulted in the greatest increase in phosphorylated ERK1 and -2 (Fig. 6B).

FIG. 6.

Detection of phosphorylated MAPK/ERK kinase in cells treated with AOP-RANTES and effect on HIV-1 replication. (A) Plot comparing the stimulatory effect of AOP-RANTES (125 nM) on HIV-1 D-92UG021 replication in PBMC pretreated with pertussis toxin or the MEK1 and -2 inhibitor, PD98059. Virus replication was measured by RT activity in the cell supernatant at day 10 and was plotted (fold) relative to the no-drug control. █, no drug; ▧, 125 nM AOP-RANTES; □, 125 nM AOP-RANTES + pertussis toxin; , 125 nM AOP-RANTES + PD98059. (B and C) Chemiluminescent Western blots showing the amounts of phosphorylated p44 ERK1 and p42 ERK2 that were immunoprecipitated from U87/CD4/CCR5 cells treated with AOP-RANTES (125 nM), HIV-1 B-92BR021, or both. For panel B, cells were treated with AOP-RANTES (125 nM) for 5 min and for 2, 5, and 24 h. Phosphorylated forms of ERK1 and -2 were immunoprecipitated from cell lysates, run on a sodium dodecyl sulfate–10% polyacrylamide gel, transferred to a nitrocellulose membrane, and probed with the same antibody specific for phosphorylated ERK1 and -2. For a positive control, TPA/PMA (50 ng/ml) was added to the U87/CD4/CCR5 cells for a strong activation of ERK1 and -2 phosphorylation. For panel C, cells were initially untreated or pretreated with pertussis toxin or PD98059 prior to exposure to virus (HIV-1 B-92BR021) alone, or virus plus 125 nM AOP-RANTES, for 5 min. Immunoprecipitations were performed with lysed cells as described above. The λ chain of the anti-ERK1 and anti-ERK2 antibodies migrated close to the p42 and p44 isoforms of ERK.

Treatment with an NSI/R5 virus, B-92BR021, also resulted in increased detection of phosphorylated MAPK/ERK (Fig. 6C), similar to the effect mediated by an SI/X4 virus or SDF-1α binding to CXCR4 (40–42). There was a slight additive effect on MAPK/ERK phosphorylation when both AOP-RANTES (125 nM) and B-92BR021 were added to cells (Fig. 6C). Pretreatment with pertussis toxin or PD98059 effectively inhibited the phosphorylation of MAPK/ERK mediated by B-92BR021 or by both B-92BR021 and AOP-RANTES (125 nM) (Fig. 6C). This finding is consistent with pertussis toxin or weak PD98059 inhibition of an AOP-RANTES-mediated stimulation of D-92UG021 HIV-1 replication in PBMC. Finally, immunoprecipitations, followed by in vitro kinase assays using glutathione-S-transferase–jun or myelin basic protein substrates, showed no activation of the stress-activated protein kinase/c-Jun N-terminal kinase by AOP-RANTES (data not shown).

AOP-RANTES treatment does not result in HIV-1 LTR activation.

Induction of the MAPK/ERK signal transduction pathway by AOP-RANTES or RANTES could lead to an increase in HIV-1 transcription from the LTR and subsequent stimulation of HIV-1 replication. To investigate this possibility, the LTR of the B-HXB2 laboratory strain, insensitive to AOP-RANTES stimulation, and the LTR from the D-92UG021 isolate, sensitive to AOP-RANTES stimulation, were cloned upstream of the luciferase gene in the pGL3 construct. These plasmids (pLTR_HXB2Luc and pLTR_D-92UG021Luc) were then transfected into three different cell lines (A549, U87/CD4/CCR5, and U87/CD4/CXCR4) prior to treatment with AOP-RANTES and RANTES (data not shown). Treatment with AOP-RANTES or RANTES did not induce transcription from the LTR of either isolate in any of the three cell lines. However, a significant induction of luciferase expression via both LTRs was observed in all cells treated with the TNF-α, a cytokine known to activate transcription factors (e.g., NF-κB) that augment transcription from the HIV-1 LTR. Based on these findings, it appears that an increase in viral transcription from the HIV-1 LTR may not be responsible for the stimulation effect by AOP-RANTES or RANTES. Poor efficiency of transfection of these LTR constructs into PBMC prevented a study of the ability of AOP-RANTES to increase transcription from the HIV-1 LTR in primary cells. However, we did perform preliminary studies to examine the effects of treatment with AOP-RANTES or RANTES on activation of NF-κB in PBMC. Using oligonucleotides derived from the NF-κB DNA binding site in the HIV-1 LTR and nuclear extracts of PBMC treated with AOP-RANTES or RANTES, we observed no significant differences in the amount of the shifted oligonucleotide–NF-κB complex in a polyacrylamide gel (data not shown). In addition, treatment with AOP-RANTES or RANTES did not result in significant activation and nuclear translocation of NF-κB in PBMC from several donors or in various cell lines expressing CCR5 (data not shown). In each case, NF-κB was activated and found in the nucleus of cells treated with TNF-α (100 ng/ml).

Simulation by AOP-RANTES and RANTES may be associated with an increase in proviral integration.

Previous reports suggest that binding of envelope glycoproteins or SDF-1α to the CXCR4 chemokine receptor can induce a signaling event (i.e., MAPK/ERK activation) and lead to an increased amount of HIV-1 proviral DNA integrated into the host cell genome (40). To test for the effects of AOP-RANTES and RANTES on integration, we PCR amplified integrated HIV-1 DNA from the same DNA samples used to amplify minus-strand strong-stop DNA (Fig. 3). The external and nested PCR amplification involved an initial external amplification with primers specific for HIV-1 DNA integrated upstream of the highly redundant Alu sequences in the host cell genome (11). This external amplification was followed by a nested amplification using primers annealing to the HIV-1 LTR. Finally, 0.18 μl of the DNA sample, from cells treated with 125 nM AOP-RANTES, was amplified with the set of nested primers (Fig. 7A). This volume represents the amount of the original DNA sample carried over from the external to nested PCR amplification. The lack of amplified DNA in this PCR confirms that products amplified from the samples by external-nested PCR were integrated copies of HIV-1 DNA (Fig. 7A).

FIG. 7.

The effect of AOP-RANTES treatment on HIV-1 integration. (A) A schematic of the HIV-1 integration assay outlines the external-nested PCR amplification technique employed to detect and quantify the amount of HIV-1 DNA integrated adjacent to Alu sequences in host DNA. The nested amplification employed a γ-³²P-labeled S2 primer and cold A2 primer to detect integrated HIV-1 DNA from the external amplification. Tenfold dilutions of pHXB2 (10³, 10⁴, and 10⁵ copies) were also PCR amplified as an amplification control. As a final control, 0.18 μl from the DNA sample treated with 125 nM AOP-RANTES was PCR amplified with the nested primer pair. This volume of sample is equivalent to the amount of the original DNA sample carried over from the external amplification (3 μl into 50 μl) into the nested PCR amplification (3 μl into 50 μl). (B) Integrated D-92UG021 DNA was quantified in PBMC treated with AOP-RANTES or RANTES. Prior to this assay, PBMC were pretreated with AOP-RANTES (0.3 or 125 nM), AOP-RANTES (125 nM) plus pertussis toxin (P.Tx.) (500 ng/ml), or RANTES (6.3 nM) and then exposed to HIV-1 D-92UG021. Cells were harvested and lysed 24 and 110 h postinfection. The external-nested PCR amplification technique (A) was then applied to detect and quantify integrated HIV-1 DNA from the samples listed above. (C) Amounts of SI/X4 HIV-1 integrated DNA (panel B) were compared to virus production 10 days postinfection. Virus production in the culture supernatant was determined using a radioactive RT assay. All values are relative to the no-drug control. (D) Integrated B-92BR021 DNA was analyzed by the same integration assay and in samples treated with the same drugs (with the exception of the AOP-RANTES + P.Tx. treatment) as described for panels A and B, respectively. (E) A plot comparing the relative amounts of NSI/R5 HIV-1 integrated DNA with relative virus production 10 days postinfection (see the legend to panel C for details).

Following a 12-h preincubation of PBMC with drug and an additional 24-h infection with the SI/X4 D-92UG021 isolates, there was a 30.9-fold increase in the amount of integrated HIV-1 DNA in samples treated with 125 nM AOP-RANTES compared to results for the no-drug control (Fig. 7B and C). Treatment with 125 nM AOP-RANTES plus pertussis toxin, RANTES, or reduced AOP-RANTES concentrations resulted in a decrease in integrated HIV-1 DNA that was still greater than that observed in the no-drug control (Fig. 7B and C). The presence of integrated HIV-1 DNA in the absence of the drug could be detected only with increased exposure of the autoradiogram (4-day versus 24-h exposure time). It is important to note that this external-nested PCR technique for detecting integrated DNA is limited in sensitivity (>10³ copies/ml) (Fig. 7A) and by the frequency of HIV-1 integration within 2 kb downstream of an Alu sequence. Given these facts, we were unable to detect a stimulation by AOP-RANTES at concentrations of less than 30 nM. The AOP-RANTES- or RANTES-mediated stimulation of HIV-1 integration at 24 h was significantly greater than the stimulation of virus production at day 10 (Fig. 7C). However, the increases of HIV-1 DNA integration at 110 h in the presence of drugs were similar to those increases in minus-strand strong-stop DNA synthesis and transcription of multiply spliced HIV-1 mRNA at 110 h as well as virus production at day 10 (see Fig. 3B and C, 4B and C, and 7C). Thus, an AOP-RANTES-mediated increase of HIV-1 integration at 24 h preempts any other increase listed above. The cells treated with AOP-RANTES and pertussis toxin had slightly less integrated DNA, suggesting that increased integration in the presence of AOP-RANTES is partially dependent on signaling through a G-protein coupled receptor. These results also suggest that the process of nuclear translocation or integration of HIV-1 DNA, and not entry or reverse transcription, is likely responsible for an AOP-RANTES-mediated stimulation of the SI/X4 D-92UG021 isolate.

Using the nested PCR amplification protocol for the detection of integrated HIV-1 DNA, we were able to amplify integrated HIV-1 DNA from samples originally exposed to an NSI/R5 isolate (B-92BR021) and treated with inhibitory concentrations of AOP-RANTES or RANTES (Fig. 7D). In contrast, minus-strand strong-stop DNA or multiply spliced mRNA could not be amplified from 24-h samples using a single-round PCR or RT-PCR protocols, respectively. Amounts of HIV-1 integrated DNA at 24 h reflected the breakthrough of virus production in the presence of 125 nM AOP-RANTES (Fig. 7D). However, we only observed a dose-dependent inhibition of HIV-1 DNA integration and virus replication with 31.5 and 0.31 nM AOP-RANTES. These results suggest that a similar mechanism involving an enhancement of HIV-1 integration may be responsible for the AOP-RANTES-mediated stimulation or breakthrough of SI/X4 or NSI/R5 HIV-1 replication, respectively.

DISCUSSION

Stimulation of HIV-1 replication by AOP-RANTES may be mediated by several mechanisms that activate various steps in the retroviral life cycle (28, 56), whereas inhibition of HIV-1 by this and other RANTES analogs occurs at the level of host cell entry (13, 51). There is at least a 100-fold difference in the range of AOP-RANTES concentrations required for the inhibition versus the stimulation, resulting in an apparent breakthrough of NSI/R5 HIV-1 replication (51, 55). Treatment with high AOP-RANTES concentrations also results in variable stimulation of SI/X4 HIV-1 replication. AOP-RANTES primes a stimulation of HIV-1 replication at the level of proviral integration by activating a signaling pathway sensitive to inhibition by pertussis toxin. Viral entry, reverse transcription, or transcription prior to integration was not affected by AOP-RANTES treatment. In addition, neither CCR5 or CXCR4 mRNA expression nor cell surface receptor expression was upregulated significantly enough by AOP-RANTES treatment to affect HIV-1 entry. Surprisingly, 125 nM AOP-RANTES mediated a significant breakthrough of NSI/R5 HIV-1 replication even though the CCR5 coreceptor could not be detected on the cell surface by FACS analysis. Recent studies have suggested that AOP-RANTES-CCR5 complexes are internalized and recycled to the cell surface (50). Continual recycling may amplify the signal transduction pathway initiated by the AOP-RANTES-CCR5 interaction as well as permitting competitive access of both virus and drug for the receptor.

Previous studies have identified a stimulation of HIV-1 replication by various β-chemokines (RANTES, MIP-1α, and MCP-3) and SDF-1 in primary cells (19, 28, 33, 36, 39, 47, 56, 58). Differential stimulation by chemokines has been attributed to the host (55–57). Variable levels of chemokine receptor expression (37, 38) or genetic polymorphisms in these genes (17, 53) could affect both inhibition and stimulation of HIV-1 replication by these β-chemokines. In addition to these host effects, we have shown that HIV-1 heterogeneity can affect the sensitivity to AOP-RANTES stimulation in the same PBMC cultures (55). Lower AOP-RANTES concentrations (100 versus 1,000 nM), comparable to those used in previous studies (28, 56), could effectively stimulate HIV-1 replication in our experiments. This may be attributable to the use of primary HIV-1 isolates and PBMC as opposed to laboratory strains and cell lines. Little or no stimulation of HIV-1 B-HXB2 was observed below a 500 nM concentration of AOP-RANTES or RANTES in CD4⁺ CCR5⁺ cell lines (28). It is possible that the introduction of exogenous CCR5 by transfection into a cell line may result in both weak CCR5 ligand signaling and subsequent stimulation of HIV-1 replication. Although AOP-RANTES did not stimulate the replication of laboratory isolates (SI/X4 B-HXB2 or NSI/R5 B-BaL) in our experiments, the addition of 125 nM AOP-RANTES to PBMC did induce a significant stimulation or breakthrough in the replication of primary SI/X4 or NSI/R5 isolates, respectively. A weak stimulation of laboratory strains compared to primary isolates may be due to reduced dependency on cell signaling pathways for virus replication and/or an adaptation to tumor cells. Finally, variable sensitivity to AOP-RANTES stimulation among primary HIV-1 isolates may be due simply to heterogeneity in the viral proteins or in the events indirectly affected by AOP-RANTES treatment. In contrast, variable sensitivity of the same primary HIV-1 isolate to AOP-RANTES stimulation in different PBMC is likely due to variable expression of RANTES-binding receptors (37, 38).

Multimerization of RANTES or AOP-RANTES (at concentrations of ≥500 nM) on glycosoaminoglycans at the cell surface appears to increase entry of B-HXB2 into host cells, as well as stimulating replication via a pertussis toxin-insensitive signaling pathway (56). In this study, stimulation of primary NSI/R5 and SI/X4 HIV-1 isolates in PBMC, by lower AOP-RANTES concentrations (≤125 nM), is diminished by pertussis toxin and was not associated with an increase in HIV entry or chemokine receptor expression. This observation highlights two potential pathways for AOP-RANTES stimulation: one specific for primary HIV-1 isolates and induced at lower concentrations (≤125 nM) and the other primarily affecting HIV-1 entry at higher concentrations (∼800 nM) (28, 56). Trkola et al. (56) have focused on the latter but have acknowledged the possibility of the former. The exact mechanism of either stimulation effect by AOP-RANTES is not well understood; however, this study suggests that an induction of a MAPK/ERK pathway by AOP-RANTES (≤125 nM) may lead to an increase in HIV-1 integration. The necessity of AOP-RANTES pretreatment for the stimulation of HIV-1 replication is likely due to an induction of a G-protein coupled signaling event, such as activation of MAPK/ERK. A signal transduction pathway can be activated via binding of HIV-1 to the G-protein coupled receptors CCR5 or CXCR4 and the CD4 receptor associated with p56_lck (6, 40).

In a series of recent studies the activation of the MAPK/ERK pathway was implicated in successful HIV-1 infection (40–42). SI/X4 HIV-1 replication in T lymphocytes requires cell activation (42), the details of which are still unknown. T-cell activation through TCR/CD28 can induce a MAPK/ERK signaling cascade (42), also shown to potentially upregulate several steps in the HIV-1 life cycle (41). In contrast, NSI/R5 viruses appear to replicate more efficiently than SI/X4 viruses in nonactivated cells (42, 59). This difference has been attributed to activation of the MAPK/ERK signaling cascade that is evidently required for SI/X4 HIV-1 replication but not for NSI/R5 replication (42). In contrast to SI/X4 HIV-1 isolates, NSI/R5 viruses were able to replicate in the presence of MEK/ERK inhibitors (42). There was no suggestion, however, that this pathway does not stimulate replication of NSI/R5 HIV-1 isolates. Another study has clearly indicated that SDF-1, recombinant gp120 derived from an SI/X4 laboratory strain, and intact SI/X4 HIV-1 strains can bind to CXCR4 and induce the MAPK/ERK pathway (40). Thus, it was not surprising that we observed a similar activation of the MAPK/ERK pathway upon binding of CCR5 to AOP-RANTES or to an NSI/R5 HIV-1 isolate. Considering that both SDF-1α and AOP-RANTES bind to different chemokine receptors (i.e., CXCR4 and CCR5, respectively) but still activate the MAPK/ERK kinase pathway, it is conceivable that the stimulation of SI/X4 HIV-1 replication by other chemokines (19, 33, 39, 47, 58) may also be mediated by similar signaling cascades following interactions with the same or different chemokine receptors. The complex signaling cascades of many seven-transmembrane coupled G-protein receptors are still being defined. This study addresses only signal transduction through the RANTES receptors and in the context of activation of HIV-1 replication.

The role of MAPK/ERK in the HIV-1 life cycle has been the subject of several recent studies. There is evidence that MAPK/ERK is able to phosphorylate HIV-1 proteins in vitro (i.e., Vif, Rev, Tat, p17^Gag, and Nef) (63, 64). However, phosphorylation of these HIV-1 proteins in vitro may not correspond to phosphorylation in vivo or have an effect on HIV-1 replication. HIV-1 Vif has effects on several steps in the HIV-1 life cycle, including viral RNA packaging and reverse transcription (27, 54), but the role of phosphorylation is unclear. Although the MAPK/ERK signaling pathway has been shown to activate several nuclear transcription factors (12), few of these have been directly implicated in transcriptional activation from the HIV-1 LTR. In our experiments, AOP-RANTES or RANTES failed to increase luciferase expression driven by the HIV-1 LTR of either B-HXB2 or D-92UG021.

The switch from a reverse transcription complex to a nuclear translocation complex during HIV-1 replication appears to be regulated by the MAPK/ERK signal transduction pathway (31). A direct interaction of phosphorylated MAPK/ERK with viral proteins in this complex may phosphorylate HIV-1 matrix protein, a step necessary for translocation of the preintegration complex to the nucleus (8). The inability of SI/X4 viruses to replicate in unstimulated T cells was related to low levels of phosphorylated MAPK/ERK and a lack of successful integration (42). In our studies, treatment with AOP-RANTES resulted in a pertussis toxin-sensitive increase in proviral DNA integration at 24 h post-HIV exposure, which diminished over time. In the presence of AOP-RANTES, increased HIV-1 mRNA transcription and reverse transcription occurred only after the first cycle of HIV-1 replication (e.g., after 24 h). In contrast, increased integration of HIV-1 in the presence of AOP-RANTES was observed as early as 24 h postinfection and during the first cycle of replication. Obviously, AOP-RANTES is not activating an event that is absolutely necessary for integration, but it may prime the cell for early nuclear translocation of the HIV-1 preintegration complex. This is consistent with the diminished stimulation by AOP-RANTES of HIV-1 integration over time (24 versus 110 h). The same signaling cascade is activated by a virus-coreceptor interaction during HIV-1 entry (6), but pretreatment with AOP-RANTES may prevent any delay and potentially enhance the association of phosphorylated MAPK/ERK with the preintegration complex (40–42).

The observation that AOP-RANTES could not stimulate laboratory isolates may be related to an adaptation of these strains to replication in tumor cells and decreased dependency on the MAPK/ERK pathway. Alternatively, the MAPK/ERK pathway may be hyperactivated or constitutively activated in many T-cell lines. However, it is important to note that a direct relationship between stimulation of virus replication, activation of MAPK/ERK, and increased HIV-1 integration is difficult to establish considering the variability of different primary HIV-1 isolates in regard to AOP-RANTES-mediated stimulation or breakthrough. Sequence analysis of the HIV-1 genes (e.g., integrase, nucleocapsid, and matrix coding regions) from these primary HIV-1 isolates may highlight specific sequence variations that may have an effect on (i) nuclear translocation of the preintegration complex and (ii) possible interactions with MAPK/ERK. This may also explain why some primary isolates show increased sensitivity to AOP-RANTES stimulation or breakthrough.

AOP-RANTES, unlike the native RANTES and other β-chemokines, does not activate the signaling cascades necessary for chemotaxis (45, 51). When taken into consideration, this observation and the increased inhibition of HIV-1 by AOP-RANTES, compared to results with RANTES, show that analogs such as AOP-RANTES are strong candidates for preclinical development and for use as drugs in HIV-1 therapy. However, members of our group have previously described variable sensitivity of primary NSI/R5 isolates to inhibition by AOP-RANTES (55). In this study, it is clear that HIV-1 replication can be stimulated by moderate AOP-RANTES concentrations (≤125 nM) by mechanisms distinct from the stimulation induced by even higher concentrations of this compound (>500 nM) (28, 56). Pretreatment of PBMC with AOP-RANTES can result in a breakthrough of primary NSI/R5 HIV-1 replication from inhibition or a stimulation of SI/X4 isolates. We have evidence that AOP-RANTES activates HIV-1 replication by increasing proviral integration through an induction of the MAPK/ERK signaling pathway.