Induction of the Gα_q Signaling Cascade by the Human Immunodeficiency Virus Envelope Is Required for Virus Entry

Abstract

Binding of human immunodeficiency virus type 1 (HIV-1) envelope glycoprotein (Env) with the primary receptor CD4 and one of two coreceptors, CXCR4 or CCR5, activates a signaling cascade resulting in Rac-1 GTPase activation and stimulation of actin cytoskeletal reorganizations critical for HIV-1-mediated membrane fusion. The mechanism by which HIV-1 Env induces Rac-1 activation and subsequent actin cytoskeleton rearrangement is unknown. In this study, we show that Env-mediated Rac-1 activation is dependent on the activation of Gα_q and its downstream targets. Fusion and Rac-1 activation are mediated by Gα_q and phospholipase C (PLC), as shown by attenuation of fusion and Rac-1 activation in cells either expressing small interfering RNA (siRNA) targeting Gα_q or treated with the PLC inhibitor U73122. Rac-1 activation and fusion were also blocked by multiple protein kinase C inhibitors, by inhibitors of intracellular Ca²⁺ release, by Pyk2-targeted siRNA, and by the Ras inhibitor S-trans,trans-farnesylthiosalicylic acid (FTS). Fusion was blocked without altering cell viability or cell surface localization of CD4 and CCR5. Similar results were obtained when cell fusion was induced by Env expressed on viral and cellular membranes and when cell lines or primary cells were the target. Treatment with inhibitors and siRNA specific for Gα_i or Gα_s signaling mediators had no effect on Env-mediated Rac-1 activation or cell fusion, indicating that the Gα_q pathway alone is responsible. These results could provide a new focus for therapeutic intervention with drugs targeting host signaling mediators rather than viral molecules, a strategy which is less likely to result in resistance.

Entry of human immunodeficiency virus type 1 (HIV-1) is mediated by sequential binding of the trimeric HIV envelope glycoprotein (Env) to CD4 and one of two primary chemokine coreceptors, CXCR4 or CCR5. This binding triggers a series of conformational changes in Env that expose the fusion peptide, which then induces the merger of viral or infected cell membranes with target membranes (14, 24, 53). The HIV-1 Env interaction with CD4 and a coreceptor also stimulates various intracellular signaling events similar to those initiated by their natural ligands, such as phosphorylation of Pyk2, Ca²⁺ mobilization, activation of RhoGTPases, and actin cytoskeleton rearrangements (15, 20, 36, 37, 48, 51, 53). Actin cytoskeletal remodeling and RhoGTPases play a central role in regulating fusion of biological membranes (13, 22). In the case of HIV-1-induced membrane fusion, activation of the RhoGTPase Rac-1 and subsequent actin cytoskeletal reorganizations are required for efficient virus entry and infection (29, 48). The exact mechanism of Env-induced Rac-1 activation that mediates actin cytoskeletal rearrangements and induces membrane fusion has not been investigated.

Binding of both CD4 and the coreceptors elicits signaling pathways that result in Rac-1 activation. However, previous results suggest that Env-induced Rac-1 activation is mediated via the coreceptor rather than CD4 (48). The coreceptor CCR5 is the principal receptor for HIV-1 transmission, whereas CXCR4 binding viral isolates are found primarily in late stages of disease (9, 17). Coreceptors CCR5 and CXCR4 belong to a family of seven-transmembrane-spanning receptors termed G protein-coupled receptors (GPCRs). Ligand-induced conformational changes in GPCRs leads to activation of heterotrimeric (het) G proteins (12, 27, 45, 53). GPCRs associate with four classes of het G proteins: Gα_i, which is sensitive to ADP ribosylation by pertussis toxin (PTX), Gα_s, Gα_q, and Gα_12/13 (12, 53). Most documented signaling through chemokine receptors goes through the PTX-sensitive Gα_i; however, fusion has been shown to be PTX insensitive (2, 6, 18, 23, 25). Additional studies have shown that mutation of the DRY domain of the second intracellular loop of CCR5, the domain thought to interact with Gα_i, also had no effect on HIV-1-induced membrane fusion and entry (2, 3, 6, 18, 23, 25). These results suggest that Env-induced fusion is independent of signaling pathways mediated by Gα_i. However, GPCRs can induce PTX-insensitive pathways by activating other het G proteins; by signaling independently of het G proteins through interactions with β-arrestin, a scaffolding protein involved in internalization of GPCRs; or by direct binding with the PDZ domain of guanine nucleotide exchange factors (GEFs) and various other binding partners (12, 28).

The signaling method utilized depends on multiple factors, including cell type, receptor, activation state of the cell, and availability of signaling partners. In addition to Gα_i, CCR5 has been shown to couple to Gα_s and Gα_q, and the interaction with Gα_q has been mapped to the third intracellular loop of CC chemokine receptors (5, 40, 53). Signaling through Gα_s leads to activation of adenylyl cyclase, calcium (Ca²⁺) channels, and cyclic AMP (cAMP)-dependent protein kinase A (PKA), whereas signaling though Gα_q results in activation of phospholipase Cβ (PLCβ), Ca²⁺ channels, and protein kinase C (PKC). Signaling components that participate in both Gα_s- and Gα_q-mediated pathways and het G protein-independent pathways are activated by HIV-1 Env interaction with CCR5 (19, 32, 36, 41). Since the likely domains involved in G protein-dependent and -independent signaling pathways are also required for normal surface expression, generation of CCR5 molecules unable to couple with signaling intermediates was not possible. In addition, similar signaling pathways are activated via CD4 and CCR5, suggesting that required signaling may occur through one receptor if the other is impaired (15, 26, 53). In this study, we utilized small interfering RNA (siRNA) and various small-molecule inhibitors to determine which signaling processes are required for Rac-1 activation and subsequent membrane fusion (Fig. 1). The data presented demonstrate that HIV-1 Env mediates activation of the Gα_q pathway via CCR5 and that this activation is critical for HIV-1-induced cell-cell fusion.

FIG. 1.

Model of signal transduction pathway elicited by Env interaction with CCR5 required for Rac-1 activation and membrane fusion. This study establishes the roles of various signaling molecules during HIV-1-induced membrane fusion by examining the effects of selective inhibitors and siRNA targeting the Gαq pathway using Rac-1 activation, virus entry, and infection assays. Signaling molecules previously shown to be activated by gp120 are shown in bold, and confirmed pathways are shown with a solid line (32, 35, 36, 48, 56). Suspected factors that may be involved are shown in lightface font, and potential pathways are shown with dotted lines. Ras is shown in black because it was shown to be activated by Env in this study. Open boxes represent siRNA or selective small-molecule inhibitors used to inhibit Gα_q and molecules downstream of Gα_q.

MATERIALS AND METHODS

Cell lines.

U87.CD4.CCR5 is an astroglioma cell line engineered to express CD4 and a CCR5 construct that is tagged at its C terminus with green fluorescent protein (GFP). The maintenance of BSC40 cells (African green monkey kidney cells), U87.CD4.CCR5 cell lines, and peripheral blood lymphocytes (PBLs) has been described previously (47). The TZM-BL (also known as JC53-BL) reporter cell line was a gift from John C. Kappes. This is a HeLa cell clone that expresses CD4, CCR5, and CXCR4 and was engineered to express Escherichia coli β-galactosidase and firefly luciferase under the transcriptional control of the HIV-1 long terminal repeat, as described previously (58). TZM-BL cells were maintained in Dulbecco's modified Eagle's medium containing 4 mM l-glutamine, 1 mM sodium pyruvate, 100 U of penicillin per ml, and 100 μg of streptomycin per ml (complete Dulbecco's modified Eagle's medium; Mediatech), supplemented with 15% fetal bovine serum (FCIII). Unless noted, tissue culture supplies were obtained from Mediatech, Manassas, VA.

Reagents.

The control-siRNA constructs (nontargeting 20- to 25-nucleotide siRNA designed as a negative control); the siRNA constructs targeted to Gα_q, Gα_i, Gα_s, Pyk2, and Rac-1; and the rabbit anti-Gα_q/11, rabbit anti-Gα_i, goat anti-Gα_s, goat anti-Pyk2, goat antiactin, and horseradish peroxidase-conjugated donkey anti-goat antibodies were obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA,). The siRNA constructs were transfected using GeneEraser siRNA transfection reagent according to the manufacturer's instructions (Stratagene, La Jolla, CA). Rabbit anti-Pyk2 antibody and thapsigargin (TG) were obtained from Sigma (St. Louis, MO). Monoclonal anti-human CCR5 antibody (MAB182) was obtained from R&D Systems (Minneapolis, MN). Rabbit polyclonal anti-Pyk2 (pY579/580) phosphospecific antibody and cytochalasin D were obtained from Invitrogen (Carlsbad, CA). GDP (100×), GTPγS (100×), and anti-Rac and anti-Ras antibodies were obtained from Millipore (Billerica, MA). U73122 was obtained from Tocris Bioscience (Ellisville, MO), U73343 and cyclopiazonic acid (CPA) were obtained from BioMol International (Plymouth Meeting, PA), xestospongin C (XC) was obtained from Cayman Chemical (Ann Arbor, MI), dantrolene was obtained from Axxora (San Diego, CA), and all other inhibitors were obtained from EMD Chemicals (San Diego, CA).

Viruses.

Wild-type vaccinia virus (WR strain) and recombinant vaccinia viruses expressing β-galactosidase (vCB21R) and T7 polymerase (vPT7-3) were obtained from the AIDS Research and Reference Reagent Program (47). Vaccinia viruses encoding an uncleaved HIV (HIV_UNC) Env (vCB-16), ADA Env (vCB-39), and HXB2 Env (vSC60) were gifts from Edward Berger, and vaccinia viruses encoding the HIV envelope from the YU2 strain (vSP-5) and constitutively active Rac GTPase (vRacV12) were gifts from C. Broder and S. Wei, respectively (46). The HIV stocks used in the assays described below were prepared by the lipofection of plasmid DNA encoding full-length proviral molecular clones, which contain the Env gene of the R5 YU2 strain in the HIV_NL4-3 backbone. To produce pseudotyped HIV-1, either amphotropic murine leukemia virus (A-MLV) envelope- or vesicular stomatitis virus G (VSV-G)-expressing plasmids were cotransfected with the proviral DNA of HIV-1. Transfected 293T cell supernatants were harvested 48 h postlipofection, filtered, and assayed for p24 antigen content by enzyme-linked immunosorbent assay (47).

Envelope-dependent fusion assay.

The HIV-1 envelope-mediated fusion assay used was a modification of an assay developed by Berger's laboratory (44). The target cells (peripheral blood mononuclear cells [PBMCs] or U87.CD4.CCR5 cells) were serum starved for 36 h and then infected overnight with vCB21R or with vRacV12. Fusion partner BSC40 cells were coinfected with vPT7-3 and vaccinia virus expressing HIV-1 Env. Cells were infected overnight (multiplicity of infection of 10 for U87.CD4.CCR5 cells and BSC40 cells and multiplicity of infection of 200 for PBMCs) at 37°C in complete media. The next day, infected cells were lightly trypsinized and washed with phosphate-buffered saline (PBS) prior to mixing. Inhibitors were added at the concentrations indicated in the figure legends to U87.CD4.CCR5 cells 1 h prior to mixing and again at the time of mixing. To allow fusion, 10⁵ cells were mixed 1:1 with a fusion partner in triplicate wells and incubated for 3 h at 37°C. To account for any effect of inhibitors on vaccinia virus infection and/or on T7 polymerase function, vCB21R- and vPT7-3-coinfected cells were similarly treated with inhibitors. Concentration curves were performed with all of the inhibitors to determine the concentration that resulted in the maximum decrease in fusion without altering vaccinia virus infection or T7 polymerase activity. Cells were lysed by the addition of NP-40 to a final concentration of 1% and freeze-thawing at −20°C. The β-galactosidase activity of reaction lysates was determined using chlorophenol red-β-d-galactopyranoside (Calbiochem), and the absorbance of each sample was determined at a wavelength of 570 nm (48).

Western blot analysis.

Cell lysates were prepared using mammalian cell lysis buffer (50 mM Tris-Cl, pH 8, 5 mM EDTA, 100 mM NaCl, 0.5% Triton X-100) plus protease and phosphatase inhibitors. Fifty micrograms of lysate was separated by 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) for siRNA-transfected samples and by 10% SDS-PAGE for Pyk2 samples; transferred to polyvinylidene difluoride membranes; blocked with 0.01% Tween 20 and 5% dry milk in 150 mM NaCl, 10 mM Tris-HCl, pH 7.5 (TBS); and probed overnight with rabbit polyclonal anti-Pyk2 (pY579/580) antibody (1:1,000) or for 2 h with anti-Gα_q (1:500), anti-Gα_i (1:500), anti-Gα_s (1:500), anti-Pyk2 (1:500), or anti-Rac (1:2,000) antibody. After a wash with TBS, a horseradish peroxidase-labeled secondary antibody was added (a 1:1,000 dilution of donkey anti-goat immunoglobulin G [IgG] [Santa Cruz], a 1:2,000 dilution of goat anti-rabbit IgG [Pierce, Rockford, IL], or a 1:10,000 dilution of anti-mouse IgG [Amersham, Piscataway, NJ]). Proteins were visualized with an Alpha Innotech (San Leandro, CA) imager by using Supersignal West Femto sensitivity substrate (Pierce). The blots were then stripped for 30 min at 50°C in stripping buffer (2% SDS, 6% Tris-Cl, pH 7.5, 0.7% β-mercaptoethanol in H₂O), washed three times in TBS, blocked and probed with antiactin (1:500) or anti-Pyk2 (1:1,000) overnight, and washed, and appropriate secondary antibodies were added prior to visualization with an Alpha Innotech imager.

Rac and Ras activation assays.

Rac and Ras activation assays were performed with 2 × 10⁶ serum-starved U87.CD4.CCR5 cells or 1 × 10⁷ serum-starved U87.CD4.CCR5 cells per sample, respectively. These cells were mixed at a ratio of 1:1 with BSC40 cells which were infected with vCB-39 (HIV_ADA Env), vSC60 (HIV_HXB2 Env), or wild-type vaccinia virus (no Env), as described above. Inhibitors were added at the concentrations indicated in the figure legends to the target cells 1 h prior to mixing and again at the time of mixing, unless otherwise noted. Reaction mixtures were incubated at 37°C for 30 min and washed two times with ice-cold PBS, and cells were lysed. Lysates were snap-frozen, and, later, equal amounts of protein per sample (determined by use of a Bio-Rad protein assay; Bio-Rad Laboratories, Hercules, CA) were analyzed using a G-LISA Rac activation assay kit (Cytoskeleton, Denver, CO) or a Ras activation assay kit (Cell Biolabs, San Diego, CA) according to the manufacturer's instructions, with equal amounts of protein for column loading (Bio-Rad).

Virus-dependent fusion assay.

A virus-induced cell fusion assay was performed as previously described (46). Briefly, U87.CD4.CCR5 cells were infected with vCB21R or vPT7-3 cultured overnight, harvested by trypsin treatment, and washed with PBS, and 10⁵ cells were mixed (1:1) with DEAE-dextran (20 μg/ml) and 100 ng p24 HIV_YU2. The assay was performed in triplicate using a 96-well microtiter plate. Inhibitors were added at the concentrations indicated in the figure legends to both populations of U87.CD4.CCR5 cells 1 h prior to virus addition and again at the time of virus addition. Fusion reaction mixtures were incubated for 3 h at 37°C, and reaction products were assayed for β-galactosidase activity as described previously (48).

JC53-BL assay.

JC53-BL cells were serum starved for 12 h and then plated overnight in complete media in a 96-well plate at 1 × 10⁴ cells per well. Cells were treated for 1 h with the concentrations of inhibitors indicated in the figure legends prior to the addition of media alone or 150 ng p24 HIV_YU2 or VSV-G- or A-MLV-pseudotyped HIV in the presence of 20 μg/ml DEAE-dextran for 3 h at 37°C. After 3 h, cells were washed three times with PBS and inhibitors were added in fresh media. Following a 24-h incubation, cells were lysed and luciferase units determined. Infected wells and uninfected wells with inhibitor were compared to wells with no inhibitor.

cAMP assay.

A cAMP assay was performed with 5 × 10⁶ serum-starved U87.CD4.CCR5 cells per condition. Half of the U87.CD4.CCR5 cells were pretreated with 200 ng/ml PTX for 18 h prior to treatment with 1 μM forskolin (EMD Chemicals, Cambridge, MA) and 500 ng/ml CCL4 (PreproTech Inc., Rocky Hill, NJ) for 30 min, and the cAMP assay was performed with equal amounts of protein per sample (Bio-Rad) according to the manufacturer's instructions (Cayman Chemical, Ann Arbor, MI).

PKA assay.

A PepTag assay for nonradioactive detection of cAMP-dependent protein kinase (Promega, Madison, WI) was used with whole-cell lysates prepared from U87.CD4.CCR5 cells treated with the inhibitors indicated in Fig. 1 and 11 for 1 h. The assay was performed in the presence of inhibitors with equal amounts of protein per sample (Bio-Rad) according to the manufacturer's protocol. PKA activity was quantitated by scanning densitometry of phosphorylated peptide substrate resolved in 0.8% Tris-HCl agarose (pH 8.0).

FIG. 11.

Pretreatment with inhibitors results in specific effects on target molecules. (A) U87.CD4.CCR5 cells were treated with the PKA inhibitor H89 (10 μM), PLC inhibitor U73122 or its negative analog U73343 (10 μM), PKC inhibitor Ch Ch (50 μM), intracellular Ca²⁺ inhibitor dantrolene (100 μM), or Ras inhibitor FTS (50 μM) for 1 h. Whole-cell lysates were then analyzed for PKA activity in the presence of inhibitors. PKA was based on standard curves of densitometry of various amounts of PKA. Results are representative of two independent experiments performed in duplicate (*, P < 0.01; **, P < 0.05). (B) Western blot analysis of phosphorylated Pyk2 from lysates of U87.CD4.CCR5 cells mixed with BSC40 cells expressing no Env (lane 1) or Env from HIV-1_HXB2 (lane 3) or HIV-1_ADA (lanes 2 and 4 to 12) at 37°C for 5 min. Cells were pretreated with DMSO alone (no inhibitor [NI], lanes 1 to 3), 1 μM TAK-779 (lane 4), 3 μM U73122 (lane 5), 3 μM U73343 (lane 6), 0.5 μM calphostin C (lane 7), 50 μM Ch Ch (lane 8), 1 μM Go-6976 (lane 9), 100 μM PKC-I(20-28) (lane 10), 100 μM dantrolene (lane 11), or 50 μM FTS (lane 12) for 1 h prior to and during the 5-min incubation with Env-expressing cells. Cell lysates were resolved by 10% SDS-PAGE, transferred to a nitrocellulose membrane, and probed with anti-Pyk2 phosphospecific (pY579/580) antibody (P-Pyk2). The same blot was then stripped and reprobed with an antibody specific for total Pyk2. The fold indicated below the P-Pyk2 bands and above the total Pyk2 bands was calculated as the quotient of the densitometry signal for the P-Pyk2 band and that for total Pyk2 (shown below) and then normalized by the ratio obtained from cells treated with DMSO alone and mixed with BSC40 cells expressing no HIV-1 Env (considered 1). Blots shown are representative of three independent experiments.

Flow cytometry and confocal microscopy.

U87.CD4.CCR5 cells were treated with the concentrations of inhibitors indicated in the figure legends for 3 h. For detection of CD4 and CCR5, phycoerythrin (PE)-conjugated anti-CD4 (Q4120; Sigma) and monoclonal anti-human CCR5 (MAB182; R&D Systems, Minneapolis, MN) antibodies were used at a 1:20-fold dilution and at 10 ng/ml, respectively. Cells were detached by incubation in 5 mM EDTA (in PBS) for 5 min at 37°C. The incubation time for the primary and secondary [goat anti-mouse PE-conjugated IgG F(ab)₂; R&D Systems] antibodies was 30 min, on ice for CCR5 and at room temperature (RT) for CD4. Cells were then fixed with 1% paraformaldehyde and analyzed using a FACSCalibur flow cytometer with CellQuest software (BD Biosciences, Franklin Lakes, NJ).

For confocal microscopy analysis, U87.CD4.CCR5 cells were plated on coverslips at 2 × 10⁵ cells per well in a six-well plate. The next day, cells were treated with the concentrations of inhibitors indicated in the figure legends for 3 h and with cytochalasin D for 15 min. Cells were fixed with 4% paraformaldehyde for 30 min at RT, blocked with 5% bovine serum albumin in PBS, and stained with anti-CD4-PE for 1 h at RT (CCR5 has a C-terminal GFP tag). Cells were washed with PBS and incubated with a 1:500 dilution of To-Pro3 (Invitrogen) in permeabilization buffer (0.2% saponin, 5% bovine serum albumin in PBS) for 15 min. Cells were washed three times in PBS and mounted on slides using Gold Antifade mounting media (Invitrogen), and 10 random fields were recorded for each condition. Cells were analyzed by confocal fluorescence microscopy using a 510 Meta LSM confocal microscope (Zeiss, Thornwood, NY).

Statistical analysis.

Raw data for fusion and infectivity assays were compared using a two-tailed t test. P values were considered significant when they were <0.05 and very significant when they were <0.01.

RESULTS

HIV-1 Env induces Rac activation and membrane fusion via a PTX-insensitive and PKA-independent pathway.

In order to determine whether HIV-1 Env-dependent Rac-1 activation is mediated by Gα_i/o family members, U87.CD4.CCR5 cells were pretreated with 200 ng/ml PTX 18 h prior to incubation with BSC40 cells expressing Env from the R5 HIV_ADA strain or cells lacking Env. Exposing cells to HIV_ADA Env in the presence or absence of PTX resulted in a 2.25-fold increase in Rac-1 activation, as determined by quantitation of levels of GTP-bound Rac1 purified by Rac-specific G-LISA (Fig. 2A). Activity of PTX was confirmed by demonstrating that macrophage inflammatory protein 1β-mediated inhibition of forskolin-induced cAMP formation was sensitive to PTX (not shown). In addition, Env-dependent cell-cell fusion was unaffected by treatment with PTX (2, 6, 18, 23, 25) (not shown). These results indicate that HIV-1 stimulates Rac-1 activity and Env-dependent cell-cell fusion by a pathway that does not involve Gα_i/o proteins.

FIG. 2.

Env-mediated Rac activation and cell-cell fusion are PTX insensitive and PKA independent. (A) U87.CD4.CCR5 cells were treated with 200 ng/ml PTX for 18 h prior to and during a 30-min incubation with BSC40 cells expressing no Env (gray) or Env from HIV-1 strain ADA (black). Whole-cell lysates were then analyzed by Rac-1-specific G-LISA activation assay (average A₄₉₀ of duplicate wells [±standard deviation] is shown; data are representative of results from three similar experiments). (B) (Top) Actin-dependent cell fusion. Average fusion compared to untreated control reactions and detected by β-galactosidase activity (±standard deviation) is shown. U87.CD4.CCR5 cells were infected with vaccinia virus expressing β-galactosidase (vCB21R) overnight, and then these cells were pretreated with DMSO alone (black) or the PKA inhibitor H89 (DMSO soluble; white) or PKA-I(14-22) (grey) at the indicated concentrations (μM) for 1 h. A portion of these cells was mixed 1:1 with HIV_ADA or HIV_UNC Env-expressing cells (subtracted as background) for 3 h in the presence of inhibitors. Data are representative of results from three similar experiments performed in triplicate. (Bottom) Cell fusion was normalized using untreated cells incubated with HIV_ADA Env as 100%. The rest of the cells were lysed, and whole-cell lysates were analyzed for PKA activity in the presence of inhibitors. PKA was based on the standard curve of A₅₇₀ of various amounts of PKA. Results are representative of three independent experiments performed in duplicate. (C) U87.CD4.CCR5 cells were treated for 1 h with 1 μM TAK-779 or 10 μM H89 prior to and during a 30-min incubation with BSC40 cells expressing no Env (subtracted as background), HIV_ADA Env (open bars), or HIV_HXB2 Env (filled bars). Whole-cell lysates were then analyzed by Rac-1-specific G-LISA activation assay. Average A₄₉₀ of duplicate wells (±standard deviation) is shown; data are representative of results from three similar experiments (*, P < 0.01).

Next, we investigated the importance of PTX-resistant G proteins Gα_s and Gα_q. The major downstream effector of Gα_s is adenylyl cyclase, which results in increased cellular levels of cAMP and leads to activation of cAMP-dependent PKA. PKA is a likely candidate for the pathway described in this study, because its activity is stimulated by HIV-1 Env and because PKA has been shown to be upstream of Rac-1 activation and subsequent cell migration (41). To test the role of PKA, U87.CD4.CCR5 cells were pretreated with two different inhibitors of PKA, inhibitory peptide PKA-I(14-22) and chemical inhibitor H89, and the effects were studied in a quantitative Env-dependent cell-cell fusion assay. Although these PKA inhibitors effectively inhibited PKA activity, there was no effect on Env-mediated cell fusion (Fig. 2B). These data confirm a previous report that synthesis of an early product of reverse transcription is not affected by PKA inhibition (4). In this study, none of the inhibitors had an effect on vaccinia virus infection or on T7 polymerase activity at the concentrations described (not shown). In addition, H89 had no effect on Env-dependent Rac-1 activation (Fig. 2C). In this experiment, controls included a mismatched X4 Env that did not induce Rac-1 activation in U87.CD4.CCR5 cells and a CCR5 inhibitor, TAK-779, that completely inhibited Rac-1 activation (48). These data suggest that although the cAMP-dependent PKA pathway is activated by HIV-1 Env, it is not required for Env-dependent Rac-1 activation and cell-cell fusion.

HIV-1 Env mediates Rac activation and membrane fusion via G proteins from the Gα_q/11 family.

In addition to Gα_s, members of the Gα_q/11 family of PTX-resistant G proteins are known to associate with CCR5 (45, 53). To investigate the importance of the Gα_q/11 family of G proteins, Gα_q expression was downregulated by RNA interference (RNAi) in U87.CD4.CCR5 cells. Gα_i and Gα_s were also downregulated to confirm that these proteins do not play a role in Env-mediated fusion and to show that knockdown of irrelevant pathways has no effect on Env-mediated fusion. siRNA-expressing cells were then mixed with BSC40 cells expressing different Env subtypes. Maximal Rac-1 activation was measured after 30 min, whereas maximal Env-dependent fusion was measured after 3 h (48). Transfection of cells with Gα_q-directed siRNAs decreased levels of Env-dependent cell-cell fusion by an average of 73% ± 3.6% for both HIV-1 R5 and dual-tropic Env subtypes (Fig. 3A, left). There was no significant fusion observed with cells expressing the X4 Env and U87.CD4.CCR5 target cells with or without siRNA expression, as was expected. Rac-1 activation was similarly decreased in cells transfected with Gα_q-directed siRNAs (Fig. 3C), whereas siRNA directed against Gα_i and Gα_s had no effect on Env-induced cell-cell fusion or Rac-1 activation (Fig. 3A and C). The decrease in levels of fusion and Rac-1 activation correlated well with the decreased steady-state level of Gα_q, and each siRNA specifically downregulated its target protein with no effects on expression of the other Gα proteins, as detected by immunoblotting (Fig. 3B). Vaccinia virus infection and T7 polymerase activity were also measured in siRNA-transfected cells, and Env-dependent cell-cell fusion in these cells was normalized accordingly. In order to determine if Gα_q acts upstream of Rac-1 in Env-mediated cell-cell fusion, a constitutively active Rac mutant, RacV12, was used (Fig. 3A, right). RacV12 overcame the effects on Env-mediated fusion resulting from siRNA targeted to Gα_q (Fig. 3A). These results show that the effects of the siRNA are overcome by expression of a downstream signaling component. Thus, the Gα_q/11 proteins are involved in HIV-1 Env-dependent cell-cell fusion upstream of Rac-1 activation.

FIG. 3.

Gα_q downregulation with siRNA reduces Env-mediated cell-cell fusion. (A) Average fusion compared to untreated control reactions and detected by β-galactosidase activity (±standard deviation) is shown. U87.CD4.CCR5 cells were transfected with 100 nM control siRNA (control) or 100 nM siRNA targeted against Gα_q, Gα_i, or Gα_s. Twenty-four hours posttransfection, U87.CD4.CCR5 cells were serum starved. Forty-eight hours posttransfection, the cells were infected with vCB21R alone or with vRacV12 in complete media. Seventy-two hours posttransfection, the transfected U87.CD4.CCR5 cells were incubated with HIV_UNC (subtracted as background), HIV_ADA, HIV_YU2, HIV_89.6, or HIV_HXB2 Env-expressing cells for 3 h. Data are representative of results from three similar experiments performed in triplicate (**, P < 0.05). Cell fusion was normalized using control-siRNA-transfected cells incubated with HIV_ADA Env as 100%. (B) Each population of transfected cells (2 × 10⁵) was lysed, and whole-cell lysates were analyzed by Western blotting. The lysates from cells transfected with 100 nM control siRNA (lane 1) or 100 nM siRNA targeted to Gα_q (lane 2), Gα_i (lane 3), or Gα_s (lane 4) were loaded on each immunoblot (IB) and were initially probed with anti-Gα_q (left), anti-Gα_i (middle), or anti-Gα_s (right); then, all blots were stripped and probed with antiactin (bottom). The relative reduction index (RI) was calculated as the quotient of the densitometry signal for the Gα_q, Gα_i, or Gα_s band and that for actin and then normalized by the ratio obtained with control siRNA (considered 1). Data represent results from one of three experiments with similar results. (C) U87.CD4.CCR5 cells were transfected with 100 nM control siRNA or 100 nM Gα_q siRNA, Gα_i siRNA, or Gα_s siRNA as described above. Seventy-two hours posttransfection, cells were incubated for 30 min with BSC40 cells expressing no Env (subtracted as background) or HIV_ADA Env. Whole-cell lysates were then analyzed by Rac-1-specific G-LISA activation assay. Average A₄₉₀ of duplicate wells (±standard deviation) is shown; data are representative of results from three similar experiments (**, P < 0.05).

HIV-1 Env-induced Rac activation and cell-cell fusion depend on PLC activation.

Activation of Gα_q leads to activation of inositol phospholipid-specific PLCβ, which catalyzes the hydrolysis of phosphatidylinositol-4,5-bisphosphate to produce the secondary messenger molecules diacylglycerol (DAG) and inositol 1,4,5-trisphosphate (IP3) (59). To test for the involvement of PLC in HIV-1-induced fusion and Rac-1 activation, U87.CD4.CCR5 cells were pretreated for 1 h with 3 μM of the PLC inhibitor U73122 or the inactive analog U73343 prior to mixture with HIV-1 Env-expressing cells and measurement of cell-cell fusion and Rac-1 activation (Fig. 4). This concentration of U73122 has previously been shown to be specific for PLC (55) and does not increase intracellular Ca²⁺, whereas higher concentrations of U73122 inhibit PLC and also lead to a release of Ca²⁺ from intracellular stores (30). The PLC inhibitor U73122 suppressed Env-dependent cell-cell fusion by an average of 73% ± 7.6% (Fig. 4A, left) and Rac-1 activation by 62% ± 1.5% (Fig. 4B). The same concentration of the inactive analog U73343 had no significant effect on fusion or Rac-1 activation (Fig. 4). Inhibitory effects of U73122 were overcome by expression of constitutively active RacV12 (Fig. 4A, right). The phosphatidylinositol-specific PLC inhibitor ET-18-OCH₃ also suppressed Env-dependent cell-cell fusion and Rac-1 activation by approximately 45% (data not shown).

FIG. 4.

PLCβ is required for Env-mediated fusion upstream of Rac-1. Average fusion compared to untreated control reactions and detected by β-galactosidase activity (±standard deviation) is shown. (A) Serum-starved U87.CD4.CCR5 cells were infected with vCB21R alone or with vRacV12 in complete media overnight. These cells were then pretreated with DMSO alone or 3 μM of PLCβ inhibitor U73122 or its negative analog U73343 for 1 h, and the inhibitors were also present during the incubation with HIV_UNC (subtracted as background), HIV_ADA, HIV_YU2, HIV_89.6, or HIV_HXB2 Env-expressing cells for 3 h. Data are representative of results from three similar experiments performed in triplicate (*, P < 0.01; **, P < 0.05). Cell fusion was normalized using DMSO-treated cells incubated with HIV_ADA Env as 100%. (B) U87.CD4.CCR5 cells were treated for 1 h with 1 μM TAK-779, 3 μM U73122, 3 μM U73343, or DMSO alone (no inhibitor) prior to and during a 30-min incubation with BSC40 cells expressing no Env (subtracted as background) or HIV_ADA Env. Whole-cell lysates were then analyzed by Rac-1-specific G-LISA activation assay. Average A₄₉₀ of duplicate wells (±standard deviation) is shown; data are representative of results from three similar experiments (*, P < 0.01; **, P < 0.05). (C) Serum-starved PBLs were infected with vCB21R in complete media overnight. These cells were then pretreated with DMSO alone, 1 μM TAK-779, or 3 μM of PLCβ inhibitor U73122 or its negative analog U73343 for 1 h, and the inhibitors were also present during the incubation with HIV_UNC (subtracted as background), HIV_ADA, HIV_YU2, HIV_89.6, or HIV_HXB2 Env-expressing cells for 3 h. Data are representative of results from three similar experiments performed in triplicate (*, P < 0.01; **, P < 0.05). Cell fusion was normalized using DMSO-treated cells incubated with HIV_ADA Env as 100%.

To validate the effects of U73122 and U73343 on Env-mediated fusion in a relevant HIV-1 target cell, PBLs which express both CCR5 and CXCR4 were used as the target cells in a quantitative Env-dependent cell-cell fusion assay. PBLs were pretreated with U73122, U73343, and TAK-779 for 1 h prior to mixture with BSC40 cells expressing different HIV-1 Env subtypes and measurement of cell-cell fusion (30). The PLC inhibitor U73122 suppressed Env-dependent cell-cell fusion in PBLs by an average of 74% ± 12.8%, and U73343 had no effect (Fig. 4C). U73122 inhibited fusion mediated by HIV-1 R5 Envs, dual-tropic Envs, and X4 Envs, suggesting that PLC is required for Env-induced fusion mediated by both CCR5 and CXCR4. TAK-779, as expected, completely inhibited fusion mediated by HIV-1 R5 Env-expressing cells but inhibited fusion mediated by the dual-tropic Env by only 59% ± 5.4% and had no significant effect on fusion mediated by X4 Env.

Intracellular Ca²⁺ release is required for Env-dependent cell-cell fusion and Rac-1 activation.

The secondary messenger IP3, produced by hydrolysis of phosphatidylinositol-4,5-bisphosphate, activates the IP3 receptor (IP3R) on the membrane of the sarcoplasmic/endoplasmic reticulum, which opens a Ca²⁺ channel resulting in intracellular Ca²⁺ release. This Ca²⁺ release leads to activation of the ryodine receptor (RyR)-operated Ca²⁺ channel, leading to a further increase in intracellular Ca²⁺. It has been shown previously that HIV-1 Env interaction with chemokine receptors results in a PTX-insensitive elevation of intracellular Ca²⁺ (15). However, the role for intracellular Ca²⁺ release in HIV-1 Env-mediated fusion is unknown. To test for the involvement of intracellular Ca²⁺ release, multiple inhibitors of intracellular calcium release were utilized. These inhibitors include XC, which specifically blocks IP3R-induced Ca²⁺ release, and dantrolene, which blocks RyR-induced intracellular Ca²⁺ release (49). In addition, TG and CPA, irreversible and reversible inhibitors, respectively, of the sarcoplasmic/endoplasmic reticulum Ca²⁺ ATPase (SERCA) pumps, were used to deplete internal Ca²⁺ stores. Inhibition of the SERCA pumps prevents Ca²⁺ from being sequestered in the endoplasmic reticulum, and Ca²⁺ is gradually lost from the cell via passive leakage into the cytosol and subsequent extrusion through the plasma membrane. TG and CPA are structurally different and are unlikely to elicit similar cellular responses unless specifically acting on SERCAs (57).

Target cells were pretreated for 1 h with 100 μM dantrolene, 10 μM CPA, 2 μM TG, or 5 μM XC. The cells were subsequently incubated with HIV-1 Env-expressing cells, and cell-cell fusion and Rac-1 activation were measured. Dantrolene, CPA, TG, and XC treatments reduced Env-mediated cell-cell fusion by averages of 93% ± 2.2%, 79% ± 2.5%, 97% ± 2.6%, and 76% ± 3.5%, respectively, in U87.CD4.CCR5 cells, which is similar to the inhibition observed for PBLs (Fig. 5A, left, and C). In the case of PBLs, this inhibition was observed with both CCR5- and CXCR4-tropic Envs. The effects of these intracellular Ca²⁺ modulators on Env-induced Rac activation in U87.CD4.CCR5 cells also correlated with their effects on Env-induced cell-cell fusion (Fig. 5B). Although all of these intracellular Ca²⁺ modifiers inhibit Env-dependent Rac-1 activation and cell-cell fusion to similar degrees, the effects on cell-cell fusion were not completely reversed by the expression of RacV12 (Fig. 5A, right). Levels of Env-dependent cell-cell fusion in cells treated with dantrolene, CPA, TG, and XC and expressing RacV12 were 50% ± 1.6%, 40% ± 2.7%, 25% ± 2.6%, and 32% ± 6% that of untreated cells, respectively, suggesting a role for intracellular Ca²⁺ flux both upstream and downstream of Rac-1 activation.

FIG. 5.

Intracellular Ca²⁺ release is required for Env-mediated fusion upstream and downstream of Rac-1. Average fusion compared to untreated control reactions and detected by β-galactosidase activity (±standard deviation) is shown. (A) Serum-starved U87.CD4.CCR5 cells were infected with vCB21R alone or with vRacV12 in complete media overnight. These cells were then pretreated with DMSO alone or 100 μM dantrolene, 10 μM CPA, 2 μM TG, or 5 μM XC for 1 h, and the inhibitors were also present during the incubation with HIV_UNC (subtracted as background), HIV_ADA, HIV_YU2, HIV_89.6, or HIV_HXB2 Env-expressing cells for 3 h. Data are representative of results from three similar experiments performed in triplicate (*, P < 0.01). Cell fusion was normalized using DMSO-treated cells incubated with HIV_ADA Env as 100%. (B) U87.CD4.CCR5 cells were treated for 1 h with DMSO alone, 1 μM TAK-779, 100 μM dantrolene, 10 μM CPA, 2 μM TG, 5 μM XC, or 100 μM RacGEF inhibitor prior to and during a 30-min incubation with BSC40 cells expressing no Env (subtracted as background) or Env from HIV_ADA Env. Whole-cell lysates were then analyzed by Rac-1-specific G-LISA activation assay. Average A₄₉₀ of duplicate wells (±standard deviation) is shown; data are representative of results from three similar experiments (*, P < 0.01). (C) Serum-starved PBLs were infected with vCB21R in complete media overnight. These cells were then pretreated with DMSO alone, 100 μM dantrolene, 10 μM CPA, 2 μM TG, or 5 μM XC for 1 h, and the inhibitors were also present during the incubation with HIV_UNC (subtracted as background), HIV_ADA, HIV_YU2, HIV_89.6, or HIV_HXB2 Env-expressing cells for 3 h. Data are representative of results from three similar experiments performed in triplicate (*, P < 0.01). Cell fusion was normalized using DMSO-treated cells incubated with HIV_ADA Env as 100%.

HIV-1 Env mediates Rac activation and cell-cell fusion via a PKC-dependent pathway.

Production of the secondary messenger DAG and IP3-induced Ca²⁺ release facilitate membrane translocation and activation of the serine/threonine kinase PKC. To investigate the role of PKC, U87.CD4.CCR5 cells were pretreated for 1 h with small-molecule inhibitors of PKC, bisindolylmaleimide I (Bis I), calphostin C, and chelerythrine chloride (Ch Ch). These cells were then incubated with HIV-1 Env-expressing cells, and cell-cell fusion was measured. The PKC inhibitors calphostin C and Ch Ch reduced Env-mediated cell-cell fusion by 87% ± 2% and 89% ± 2%, respectively, whereas Bis I reduced fusion by only 55% ± 11% (Fig. 6A, left). Concentration curves were performed with all of the inhibitors used in this study to determine the concentration that resulted in the maximum decrease in fusion without altering vaccinia virus infection or T7 polymerase activity (data not shown). At higher concentrations, Bis I had a greater effect on Env-induced cell fusion; however, at these concentrations, Bis I is no longer specific for PKC (56) (data not shown). Effects of PKC inhibition on fusion were overcome with constitutively active RacV12 (Fig. 6A, right).

FIG. 6.

PKC is required for Env-mediated fusion upstream of Rac-1. Average fusion compared to untreated control reactions and detected by β-galactosidase activity (±standard deviation) is shown. (A and B) Serum-starved U87.CD4.CCR5 cells were infected with vCB21R alone or with vRacV12 overnight in complete media. These cells were then pretreated with DMSO alone or the PKC inhibitor Bis I (10 μM), calphostin C (0.5 μM), or Ch Ch (50 μM) for 1 h (A) or pretreated with the specific PKC inhibitor Go-6976 (Ca²⁺ dependent, 1 μM), Ro-32-0432 (selective for PKCα and PKCβ, 10 μM), or PKC-I(20-28) (myristolated peptide inhibitor specific for PKCα and PKCβ, 100 μM) for 1 h (B). The inhibitors were also present during the incubation with HIV_UNC (subtracted as background), HIV_ADA, HIV_YU2, HIV_89.6, or HIV_HXB2 Env-expressing cells for 3 h. Data are representative of results from three similar experiments performed in triplicate (*, P < 0.01; **, P < 0.05). Cell fusion was normalized using DMSO-treated cells incubated with HIV_ADA Env as 100%. (C) U87.CD4.CCR5 cells were treated for 1 h with DMSO alone (no inhibitor), 1 μM TAK-779, 0.5 μM calphostin C, 50 μM Ch Ch, 1 μM Go-6976, 100 μM PKC-I(20-28), or 100 μM RacGEF inhibitor prior to and during a 30-min incubation with BSC40 cells expressing no Env (subtracted as background) or HIV_ADA Env. Whole-cell lysates were then analyzed by Rac-1-specific G-LISA activation assay. Average A₄₉₀ of duplicate wells (±standard deviation) is shown; data are representative of results from three similar experiments (*, P < 0.01). (D) Serum-starved PBLs were infected with vCB21R overnight in complete media. These cells were then pretreated with DMSO alone or the PKC inhibitor Ch Ch (50 μM) or Go-6976 for 1 h. The inhibitors were also present during the incubation with HIV_UNC (subtracted as background), HIV_ADA, HIV_YU2, HIV_89.6, or HIV_HXB2 Env-expressing cells for 3 h. Data are representative of results from three similar experiments performed in triplicate (*, P < 0.01). Cell fusion was normalized using DMSO-treated cells incubated with HIV_ADA Env as 100%.

To determine which isoform of PKC is involved in Env-mediated Rac-1 activation and cell-cell fusion, we used PKC inhibitors that are specific for conventional isoforms PKCα, PKCβ, and PKCγ, which depend on Ca²⁺ and DAG as well as a phospholipid for activation. The specific PKC inhibitors used were Ro-32-0432, which is most specific for PKCα; Go-6976, which specifically inhibits Ca²⁺-dependent PKC, particularly PKCα and PKCβI; and the inhibitory peptide PKC-I(20-28), which is a pseudosubstrate that specifically blocks PKCα and PKCβ activation. U87.CD4.CCR5 cells were pretreated with these inhibitors for 1 h prior to mixture with HIV-1 Env-expressing cells and measurement of cell-cell fusion. Pretreatment with Ro-32-0432, Go-6976, and PKC-I(20-28) decreased Env-mediated cell-cell fusion by 73% ± 4.4%, 70% ± 3.6%, and 93% ± 3.5%, respectively (Fig. 6B, left). Expression of the constitutively active Rac mutant RacV12 overcame the effects of these PKC inhibitors on HIV-1 Env-mediated cell-cell fusion (Fig. 6B, right).

To determine the effects of PKC inhibitors on Rac-1 activation, U87.CD4.CCR5 cells were pretreated with calphostin C, Ch Ch, Go-6976, and PKC-I(20-28) for 1 h prior to mixture with HIV-1 Env-expressing cells. Levels of suppression of Env-dependent Rac-1 activation correlated with levels of inhibition of Env-dependent cell-cell fusion (Fig. 6C). In this experiment, controls included the CCR5 inhibitor TAK-779 and a specific RacGEF inhibitor, both of which inhibited Env-induced Rac-1 activation (44). In addition to the data with RacV12, this suggests that PKCα and/or PKCβ is required upstream of Rac-1 in Env-mediated cell fusion.

To validate the effects of both general and specific PKC inhibitors on Env-mediated fusion in a relevant HIV-1 target cell, PBLs were pretreated with Ch Ch and Go-6976 for 1 h prior to mixture with HIV-1 Env-expressing cells and measurement of cell-cell fusion. Pretreatment with Ch Ch and Go-6976 decreased Env-mediated cell-cell fusion with HIV-1 R5 Envs, dual-tropic Env, and X4 Env by averages of 98% ± 2.9% and 75% ± 2%, respectively (Fig. 6D). As with the above-described experiments with the PLC inhibitor and Ca²⁺ inhibitors, these results suggest that PKC is required for Env-induced cell-cell fusion in cell lines and primary cells regardless of Env coreceptor preference.

HIV-1 Env-induced Pyk2 activation mediated by intracellular Ca²⁺ release is required for Env-dependent cell-cell fusion and Rac-1 activation.

Previous studies have shown that HIV-1 Env-induced intracellular Ca²⁺ release results in activation of the focal-adhesion-related kinase Pyk2, a downstream target of PKC (15). Pyk2 is a nonreceptor tyrosine kinase that can be activated by growth factors, chemokines, and GPCR ligands and provides a link between these ligands and multiple downstream cellular events, such as modulation of the cytoskeleton (15, 32). There are no available inhibitors specific for Pyk2, so to test the role of Pyk2 in Env-induced Rac-1 activation and cell-cell fusion, Pyk2 expression was downregulated by RNAi in U87.CD4.CCR5 cells. RNAi to Rac-1 was used as a positive control for downregulation of fusion. siRNA-expressing cells were then mixed with BSC40 cells expressing different Env subtypes. Transfection of cells with Pyk2- and Rac-1-directed siRNAs decreased levels of Env-dependent cell-cell fusion by averages of 77% ± 5.4% and 73% ± 2.3%, respectively (Fig. 7A, left), and Rac-1 activation by 83% ± 2.5% and 81% ± 0.6%, respectively (Fig. 7C). The decrease in levels of fusion and Rac-1 activation correlated well with the decreased steady-state levels of Pyk2 and Rac-1 protein, and each siRNA was specific for its target protein, as detected by immunoblotting (Fig. 7B). Vaccinia virus infection and T7 polymerase activity were also measured in siRNA-transfected cells, and Env-dependent cell-cell fusion in these cells was normalized accordingly. RacV12 overcame the effects on Env-mediated fusion resulting from siRNA targeted to Pyk2 or Rac-1 (Fig. 7A, right), indicating that Pyk2 acts upstream of Rac-1 in Env-mediated cell-cell fusion and that the effects of the siRNA are overcome by Rac-1 overexpression.

FIG. 7.

Env-induced fusion and Rac-1 activation are dependent on Pyk2. (A) Average fusion compared to untreated control reactions and detected by β-galactosidase activity (±standard deviation) is shown. U87.CD4.CCR5 cells were transfected with 100 nM control siRNA (control) or 100 nM siRNA targeted against Pyk2 or Rac-1. Twenty-four hours posttransfection, U87.CD4.CCR5 cells were serum starved. Forty-eight hours posttransfection, the cells were infected with vCB21R alone or with vRacV12 in complete media. Seventy-two hours posttransfection, the transfected U87.CD4.CCR5 cells were incubated with HIV_UNC (subtracted as background), HIV_ADA, HIV_YU2, HIV_89.6, or HIV_HXB2 Env-expressing cells for 3 h. Data are representative of results from three similar experiments performed in triplicate (*, P < 0.01). Cell fusion was normalized using control-siRNA-transfected cells incubated with HIV_ADA Env as 100%. (B) Each population of transfected cells (2 × 10⁵) was lysed, and whole-cell lysates were analyzed by Western blotting. The lysates from cells transfected with 100 nM control siRNA (lane 1) or 100 nM siRNA targeted to Pyk2 (lane 2) or Rac-1 (lane 3) were loaded on each immunoblot (IB) and were initially probed with anti-Pyk2 (left) or anti-Rac-1 (right); then, all blots were stripped and probed with antiactin (bottom). The relative reduction index (RI) was calculated as the quotient of the densitometry signal for the Pyk2 or Rac-1 band and that for actin and then normalized by the ratio obtained with control siRNA (considered 1). Data represent results from one of three experiments with similar results. (C) U87.CD4.CCR5 cells were transfected with 100 nM control siRNA, 100 nM Pyk2 siRNA, or 100 nM Rac-1 siRNA as described above. Seventy-two hours posttransfection, cells were incubated for 30 min with BSC40 cells expressing no Env (subtracted as background) or HIV_ADA Env. Whole-cell lysates were then analyzed by Rac-1-specific G-LISA activation assay. Average A₄₉₀ of duplicate wells (±standard deviation) is shown; data are representative of results from three similar experiments (**, P < 0.05).

HIV-1 Env-induced Rac activation and cell-cell fusion depend on Ras activation.

Pyk2 can lead to Rac-1 activation via multiple signaling partners; however, previous data suggest that Env-dependent Rac-1 activation and cell-cell fusion are mediated by a Rac-specific GEF (46). The Rac-specific GEF Tiam1 is directly activated by Ras and is involved in actin cytoskeletal reorganization (35, 42). In addition, Pyk2 activates Ras via Src downstream of Gα_q, providing a link between Pyk2 and Rac-1 activation (10, 43). To test the role of Ras, U87.CD4.CCR5 cells were pretreated with the Ras inhibitor S-trans,trans-farnesylthiosalicylic acid (FTS), which blocks membrane association of Ras, thus preventing Ras activation. FTS inhibition of Env-dependent cell-cell fusion was dose dependent (Fig. 8A, left), and concentrations of FTS shown to abolish Env-dependent cell-cell fusion in U87.CD4.CCR5 cells similarly inhibited Env-dependent Rac-1 activation and Env-dependent cell-cell fusion induced by BSC40 cells expressing all HIV-1 Envs with target PBLs (Fig. 8B and C). However, FTS inhibition of cell-cell fusion was not completely reversed by expression of RacV12, suggesting a role for Ras both upstream and downstream of Rac-1 activation (Fig. 8A, right).

FIG. 8.

Ras is required for Env-mediated fusion upstream and downstream of Rac-1. Average fusion compared to untreated control reactions and detected by β-galactosidase activity (±standard deviation) is shown. (A) Serum-starved U87.CD4.CCR5 cells were infected with vCB21R alone or with vRacV12 in complete media overnight. These cells were then pretreated with DMSO alone or the indicated concentrations of the Ras inhibitor FTS for 1 h, and the inhibitor was also present during the incubation with HIV_UNC (subtracted as background), HIV_ADA, HIV_YU2, HIV_89.6, or HIV_HXB2 Env-expressing cells for 3 h. Data are representative of results from three similar experiments performed in triplicate (*, P < 0.01). Cell fusion was normalized using DMSO-treated cells incubated with HIV_ADA Env as 100%. (B) U87.CD4.CCR5 cells were treated for 1 h with DMSO alone (no inhibitor), 1 μM TAK-779, 50 μM FTS, or 100 μM RacGEF inhibitor prior to and during a 30-min incubation with BSC40 cells expressing no Env (subtracted as background) or Env from HIV_ADA Env. Whole-cell lysates were then analyzed by Rac-1-specific G-LISA activation assay. Average A₄₉₀ of duplicate wells (±standard deviation) is shown; data are representative of results from three similar experiments (*, P < 0.01; **, P < 0.05). (C) Serum-starved PBLs were infected with vCB21R in complete media overnight. These cells were then pretreated with DMSO alone, 50 μM Ras inhibitor FTS, or 100 μM RacGEF inhibitor for 1 h, and the inhibitor was also present during the incubation with HIV_UNC (subtracted as background), HIV_ADA, HIV_YU2, HIV_89.6, or HIV_HXB2 Env-expressing cells for 3 h. Data are representative of results from three similar experiments performed in triplicate (*, P < 0.01). Cell fusion was normalized using DMSO-treated cells incubated with HIV_ADA Env as 100%. (D) Ras activation assay. Western blot analysis of Raf-1 binding fractions from lysates of U87.CD4.CCR5 cells mixed with BSC40 cells expressing no Env (lane 3), HIV_HXB2 Env (lane 4), or HIV_ADA Env (lanes 5 to 9). TAK-779 (1 μM) was included to inhibit CCR5-Env binding (lane 5). Cells were pretreated with DMSO alone (lane 6) or the indicated concentrations of FTS to inhibit Ras (lanes 7 to 9) for 1 h prior to and during the 30-min incubation with Env-expressing cells. Positive (lane 1) and negative (lane 2) controls were generated by GTPγS and GDP loading of reaction lysates, respectively. Increases (n-fold) in the amount of Ras-GTP compared to that in lane 3 were determined by densitometry (normalized to lysate loading control) and are indicated. Data represent results from one of three experiments with similar results.

Ras activation downstream of the chemokine receptor CCR5 has not been shown in previous studies. To determine if Ras is activated via HIV-1 Env interaction with CCR5 and to determine if FTS inhibits this activity, U87.CD4.CCR5 cells were pretreated for 1 h with dimethyl sulfoxide (DMSO) alone or with the concentrations indicated in Fig. 8D of FTS prior to mixture with HIV-1 Env-expressing cells for 30 min and measurement of Ras activation. The activation state of Ras in cell lysates was determined by utilizing Raf-1 binding to separate Ras-GTP from total Ras (Fig. 8D). There was a 3.6-fold increase in activated Ras when U87.CD4.CCR5 cells were mixed with cells expressing Env from the R5 HIV_ADA strain (Fig. 8D, lane 6) versus when U87.CD4.CCR5 cells were mixed with cells expressing no HIV Env (Fig. 8D, lane 3) or expressing Env from the X4 HIV_HXB2 strain (Fig. 8D, lane 4). The increase in levels of activated Ras was prevented by TAK-779 (Fig. 8D, lane 5), suggesting that the Env interaction with the coreceptor, rather than with CD4 alone, mediates Ras activation. Env-induced Ras activation was inhibited by FTS in a dose-dependent manner that correlates with the inhibition of Env-dependent cell-cell fusion (Fig. 8A and D, lanes 7, 8, and 9). These data suggest that HIV-1 Env-induced Ras activation is required for Env-dependent Rac-1 activation and cell-cell fusion.

HIV-1-dependent cell fusion and infection of TZM-BL cells are dependent on the Gα_q signaling pathway.

The Env-induced cell fusion assay is a rapid, sensitive, and flexible method used to solely study the HIV-1 Env-mediated fusion step of virus infection. However, this assay involves overexpression of Env on the cell surface of BSC40 cells and does not involve virus particles and may not reflect all of the factors that mediate fusion. Therefore, we used a virus-dependent fusion assay that is based on the ability of virus particles to bridge two cells and allow transfer of cytoplasmic content (46). This assay has advantages over the Env-induced cell fusion assay since it utilizes relevant levels of virus-associated glycoprotein and is representative of virus-cell fusion versus cell-cell fusion. In this assay, we used two populations of U87.CD4.CCR5 cells: one population was infected with vaccinia virus expressing T7 polymerase, and the second population was infected with a vaccinia virus expressing the β-galactosidase gene under the regulation of the T7 promoter. Both populations of U87.CD4.CCR5 cells were treated with TAK-779, U73122, U73343, Ch Ch, Go-6976, PKC-I(20-28), dantrolene, CPA, TG, XC, or FTS for 1 h prior to the addition of fusion-competent R5 virus HIV_YU2 for 3 h. In this experiment, controls included untreated and inhibitor-treated cells that were not incubated with virus (subtracted as background) and the CCR5 inhibitor TAK-779, which completely inhibited HIV-1 virus-dependent cell fusion. Like Env-dependent fusion, virus-induced fusion was reduced by 73% ± 5% in U73122-treated cells versus cells treated with DMSO alone, and the inactive analog U73343 had no effect (Fig. 9A). The PKC inhibitors dantrolene, CPA, TG, XC, and FTS also inhibited virus-dependent fusion at levels comparable to inhibition seen with the Env-dependent fusion assay (Fig. 9A).

FIG. 9.

Effect of inhibitors on virus-dependent cell fusion with HIV-1 R5 virus and infection of TZM-BL cells with HIV-1 R5 virus or A-MLV-pseudotyped or VSV-G-pseudotyped HIV-1 virus. (A) Average fusion compared to untreated control reactions and detected by β-galactosidase activity (±standard deviation) is shown. One population of serum-starved U87.CD4.CCR5 cells was infected with vCB21R and the other population of U87.CD4.CCR5 cells was infected with vPT7-3 in complete media overnight. These cells were then mixed (1:1) in triplicate wells of a 96-well plate and pretreated with DMSO alone, TAK-779 (1 μM), U73122 (3 μM), U73343 (3 μM), Ch Ch (50 μM), Go-6976 (1 μM), PKC-I(20-28) (100 μM), dantrolene (100 μM), CPA (10 μM), TG (2 μM), XC (5 μM), or FTS (50 μM) for 1 h. Inhibitors were also present during incubation with 100 ng of HIV_YU2 per well for 3 h at 37°C. Data are representative of results from three similar experiments (*, P < 0.01). Cell fusion was normalized using DMSO-treated cells mixed with HIV_YU2 as 100%. VDFA, virus-dependent fusion assay. (B) JC53-BL cells were preincubated with DMSO alone, TAK-779 (1 μM), U73122 (3 μM), U73343 (3 μM), Ch Ch (50 μM), Go-6976 (1 μM), dantrolene (100 μM), CPA (10 μM), TG (2 μM), XC (5 μM), FTS (50 μM), NH₄Cl (50 mM), or OA (100 nM) for 1 h, and inhibitors were also present during the 3-h infection period with 150 ng of HIV_YU2, A-MLV-pseudotyped HIV-1, or VSV-G-pseudotyped HIV-1 per well. Virus and inhibitor were then washed off of the cells, and the cells were incubated in the same concentration of inhibitor at 37°C overnight. Luciferase activities in the infected cell lysates were measured 24 h postinfection and were used to calculate virus infectivity relative to that of the control. Data are representative of results from three similar experiments performed in triplicate (**, P < 0.05). Cell infection was normalized using DMSO-treated cells as 100%. (C) JC53-BL cells were transfected with 100 nM control siRNA (control) or 100 nM siRNA targeted against Gα_q, Gα_i, Gα_s, Pyk2, or Rac-1. Twenty-four hours posttransfection, JC53-BL cells were serum starved. Forty-eight hours posttransfection, the cells were infected with 150 ng of HIV_YU2 for 3 h. Virus was then washed off, and the cells were incubated at 37°C overnight. Luciferase activities in the infected cell lysates were measured at 24 h postinfection and were used to calculate virus infectivity relative to that of the control. Data are representative of results from three similar experiments performed in triplicate (**, P < 0.05). Cell infection was normalized using control siRNA-transfected calls infected with 150 ng of HIV_YU2 as 100%.

The inhibitory effects of the various inhibitors to the downstream mediators of Gα_q on HIV-1 Env-induced cell fusion and virus-dependent cell fusion suggest an inhibition of virus entry. However, both the Env-dependent and virus-dependent fusion assays measure only the initial step of virus infection and do not look at the effect of blocking the Gα_q pathway and, subsequently, fusion on the entire virus life cycle. It is also unknown whether the inhibitors used in this study specifically block fusion induced by HIV-1 Env or if they block virus-induced membrane fusion and infection in general. To test the specificity of these inhibitors for HIV-1 Env-induced entry and infection, we performed the following assay with wild-type HIV_YU2 particles or with VSV-G- and A-MLV-pseudotyped HIV-1 virus particles. HIV-1 enters cells by inducing virus-cell membrane fusion at the cell surface. When HIV-1 is pseudotyped with VSV-G or A-MLV Env, the route of viral entry becomes clathrin-mediated or caveola-mediated endocytosis (8, 39). To analyze the effects of these inhibitors on infection with each of these viruses, we performed a single-cycle HIV-1 infection assay using TZM-BL indicator cells (also known as JC53-BL cells), a derivative of HeLa cells, which express surface levels of CD4, CCR5, and CXCR4 comparable to levels produced by PBMCs (21). These cells have also been engineered to express β-galactosidase and luciferase under the control of the HIV-1 long terminal repeat, which allows quantitative measurement of virus infectivity. TZM-BL cells were pretreated with TAK-779, U73122, U73343, Ch Ch, Go-6976, dantrolene, CPA, TG, XC, or FTS for 1 h prior to incubation with R5 virus HIV_YU2, VSV-G-pseudotyped HIV-1, or A-MLV-pseudotyped HIV-1 for 3 h. Ammonium chloride (NH₄Cl) (inhibits clathrin-mediated endocytosis) and okadaic acid (OA) (inhibits caveola-mediated endocytosis) were included as controls for the inhibition of VSV-G- and A-MLV-mediated entry, respectively. After 3 h, virus was washed off, the same concentration of inhibitor was added back to the wells, and the cells were incubated at 37°C for 24 h. Figure 9B shows that the reductions in infection with HIV_YU2 in the presence of PKC inhibitors, Ca²⁺ inhibitors, and FTS were comparable to the reductions in Env-dependent and virus-dependent cell fusion (Fig. 9B, top left). The stronger inhibitory effect of U73122 seen in the reporter gene virus infectivity assay suggests that PLC could also play a role in postfusion steps in the virus life cycle (Fig. 9B, top left). Interestingly, VSV-G HIV-1 infected inhibitor-treated cells, with the exception of U73122- and NH₄Cl-treated cells, at efficiencies equal to those of DMSO-treated cells (Fig. 9B, top right). Similarly, A-MLV HIV-1 infected inhibitor-treated cells, with the exception of U73122- and OA-treated cells, as efficiently as it did DMSO-treated cells (Fig. 9B, bottom left). These data indicate that the PKC inhibitors, Ca²⁺ inhibitors, and FTS were unable to block HIV-1 infection when HIV-1 entered cells through VSV-G-mediated or A-MLV-mediated endocytosis. In other words, reverse transcription and nuclear import were able to proceed in the presence of these inhibitors. VSV-G HIV-1 and A-MLV HIV-1 infections in U73122-treated cells were inhibited by 57% ± 8.9% and 64% ± 3.6%, respectively (Fig. 9B). These results support the suggestion that PLC plays a role in postfusion steps in the virus life cycle.

To analyze the effects of the siRNAs targeted to Gα_q, Gα_i, Gα_s, Pyk2, and Rac-1 on HIV-1 infection and to show that the effects are not cell type specific, we performed a single-cycle HIV-1 infection assay using TZM-BL indicator cells transfected with control siRNA and siRNA targeted to Gα_q, Gα_i, Gα_s, Pyk2, and Rac-1. Figure 9C shows that the reduction in infection with HIV_YU2 in cells expressing siRNA directed against Gα_q, Pyk2, and Rac-1 versus that in cells expressing control siRNA was comparable to the reduction in Env-dependent cell-cell fusion and Rac-1 activation (Fig. 3 and 7), whereas the siRNA targeted to Gα_i and Gα_s had no effect. The decreased steady-state levels of Gα_q, Gα_i, Gα_s, Pyk2, and Rac-1 protein in TZM-BL cells expressing targeted siRNA versus control siRNA were similar to that observed for U87.CD4.CCR5 cells (Fig. 3B and 7B), and each siRNA was specific for its target protein, as detected by immunoblotting (data not shown).

Inhibitor treatment does not alter CD4/CCR5 receptor expression or surface localization.

Suppression of Env-dependent Rac-1 activation and cell-cell fusion in the presence of inhibitors could also be explained by a change in CD4/CCR5 receptor localization or surface expression. To confirm that there were no changes in CD4/CCR5 receptor localization, U87.CD4.CCR5 cells were incubated with the inhibitors for 3 h and were analyzed by confocal microscopy. No gross differences in cell surface localization of CD4 or CCR5 were observed (Fig. 10A). Blocking CCR5 with TAK-779 and Rac-1 activation with the RacGEF inhibitor also had no effect on CD4/CCR5 localization. In contrast, CD4 and CCR5 localization could be disrupted by treating cells with the actin assembly inhibitor cytochalasin D. U87.CD4.CCR5 cells treated with inhibitors reconstituted in DMSO were compared to DMSO-treated cells, and cells treated with inhibitors reconstituted in H₂O were compared to untreated cells.

FIG. 10.

Gα_q inhibitors do not affect surface expression and localization of CD4 and CCR5 GFP. (A) Confocal micrographs of U87.CD4.CCR5 cells untreated or treated with DMSO alone, TAK-779 (1 μM), U73122 (3 μM), U73343 (3 μM), calphostin C (0.5 μM), Ch Ch (50 μM), PKC-I(20-28) (100 μM), dantrolene (100 μM), FTS (50 μM), or RacGEF inhibitor (100 μM) for 3 h or with cytochalasin D (1 μM) for 15 min, fixed, stained with anti-CD4-PE antibody (red), and counterstained with TO-PRO3 (blue). The green GFP signal and the red PE signal have been merged to show areas of colocalization (yellow). Data are representative of results from three experiments. Images were collected using an oil objective (magnification, ×63). (B) U87.CD4.CCR5 cells were incubated for 3 h with all inhibitors listed for panel A except cytochalasin D and detached by treatment with 5 mM EDTA. U87 cells and untreated and treated U87.CD4.CCR5 cells were labeled for surface expression of CD4 and CCR5 and analyzed by flow cytometry. Unlabeled U87.CD4.CCR5 cells were used to compensate for GFP. Data are expressed as percentages of surface expression based on DMSO-treated cells as 100%. Cal C, calphostin C.

To determine the effect of inhibitors on surface expression of the CD4/CCR5 receptors, we used flow cytometric analysis. Incubation of U87.CD4.CCR5 cells with U73122, U73343, calphostin C, Ch Ch, PKC-I(20-28), dantrolene, FTS, or the RacGEF inhibitor for 3 h had no effect on cell surface expression of CD4 or CCR5 (Fig. 10B). In this experiment, U87 cells that do not express CD4 or CCR5 were used as a negative control and untreated cells were used as a positive control. TAK-779, a CCR5 inhibitor that interacts with the same domain as the anti-CCR5 antibody, blocked CCR5 staining but had no effect on CD4 staining. These results confirm that at the concentrations used in this study these inhibitors do not alter CD4/CCR5 cell surface expression or localization.

Inhibitor treatment results in specific inhibition of target molecules.

Next, we examined whether the decrease in Env-dependent Rac-1 activation and cell-cell fusion in the presence of inhibitor is completely due to inhibition of the target molecule by determining effects on PKA activity and Pyk2 phosphorylation. To measure the effect on PKA activity, U87.CD4.CCR5 cells were pretreated for 1 h with U73122, U73343, Ch Ch, dantrolene, or FTS, and then cell lysates were analyzed using a PepTag assay for the nonradioactive detection of cAMP-dependent protein. PKA activity in U73122-, U73343-, ChCh-, and FTS-treated cells was similar to that in untreated cells (Fig. 11A). PKC inhibitors Bis I, calphostin C, Go-6976, and Ro-32-0432 were also tested, with similar results (data not shown). Treatment of U87.CD4.CCR5 cells with dantrolene decreased PKA activity by 30%, suggesting a role for intracellular Ca²⁺ release in maintaining endogenous PKA activity. This decrease is minimal compared to the 75% reduction in PKA activity in cells treated with H89 (Fig. 11A). In addition, dantrolene-treated cells showed a 90% decrease in Env-induced Rac-1 activation and a 98% decrease in cell-cell fusion, whereas PKA inhibition had no effect (Fig. 2B and 5).

To determine the effect of inhibitors on HIV-1 Env-induced Pyk2 activation, U87.CD4.CCR5 cells were pretreated with U73122, U73343, calphostin C, Ch Ch, Go-6976, PKC-I(20-28), dantrolene, or FTS for 1 h prior to mixture with HIV-1 Env-expressing cells for 5 min and measurement of Pyk2 phosphorylation. As expected, inhibitors targeted to PLC, PKC, and intracellular Ca²⁺ release that are required upstream of Pyk2 decreased the amount of Env-induced Pyk2 phosphorylation, whereas FTS had no effect (Fig. 11B).

Table 1 summarizes the effects of the PLC inhibitor, general and specific PKC inhibitors, intracellular Ca²⁺ release inhibitors, and FTS on Env-induced cell-cell fusion, Rac-1 activation, virus-dependent cell-cell fusion, infection of TZM-BL cells, PKA activation, and Pyk2 phosphorylation. These results confirm that at the concentrations used in this study these inhibitors are specific for their target molecules.

TABLE 1.

Summary of studies with inhibitors^a

Reagent (μM)	Target	% EDFa	% EDFb with RacV12	% Rac-1 activation	% VDFc	% Infection	% PKA activation	% Pyk2 phosphorylation
H89 (10)	PKA	95 ± 2	NTd	100 ± 2	NT	NT	24 ± 0.6	NT
U73122 (3)	PLC	27 ± 1	108 ± 0.2	27 ± 2	28 ± 5	0.8 ± 1	105 ± 5	19
U73343 (3)	Inactive analog for U73122	95 ± 1	105 ± 0.4	104 ± 4	99 ± 14	96 ± 10	111 ± 4	97
Bis I (10)	PKC	45 ± 11	107 ± 8	NT	NT	NT	NT	NT
Calphostin C (0.5)	PKC	13 ± 2	97 ± 7	14 ± 1	NT	NT	NT	16
Ch Ch (50)	PKC	11 ± 2	116 ± 9	17 ± 0.4	7 ± 4	10 ± 2	99 ± 4	14
Go-6976 (1)	Ca2+-dependent PKC	21 ± 4	99 ± 3	27 ± 1	24 ± 6	33 ± 3	NT	26
Ro-32-0432 (10)	Specific for PKCα and PKCβ	27 ± 4	108 ± 5	NT	NT	NT	NT	NT
PKC-I(20-28) (100)	Specific for PKCα and PKCβ	7 ± 3	114 ± 5	6 ± 1	0 ± 3	2 ± 0.2	NT	0
Dantrolene (100)	RyR-intracellular Ca2+ release	7 ± 2.2	50 ± 1.6	10 ± 5	9 ± 1.8	10 ± 1.2	70 ± 2.5	0
CPA (10)	SERCA	21 ± 2.5	40 ± 1.6	19 ± 1.2	14 ± 3	13 ± 1.7	NT	NT
TG (2)	SERCA	3 ± 2.6	25 ± 1.6	21 ± 2.2	9 ± 1.1	0.6 ± 0.5	NT	NT
XC (5)	IP3R-intracellular Ca2+ release	24 ± 3.5	32 ± 1.6	25 ± 1.3	20 ± 6.6	3 ± 0.5	NT	NT
FTS (50)	Ras	0 ± 0.2	40 ± 4	20 ± 4	0 ± 9	0.3 ± 2	92 ± 24	97

^aValues are percentages determined for treated reactions compared to those determined for reactions with untreated controls.

^bEDF, Env-dependent fusion (percent fusion is the average of reactions from cells incubated with HIV_ADA, HIV_YU2, and HIV_89.6 Env-expressing cells).

^cVDF, virus-dependent fusion.

^dNT, not tested.

DISCUSSION

Earlier studies using PTX and/or mutant CD4 and coreceptors supported the view that signaling is not required for HIV-1 Env-mediated fusion (2, 6, 18, 23, 25). However, it is now apparent that signaling via CD4 and the coreceptors is far more intricate than was thought previously. The studies suggesting that signaling is dispensable for HIV-1 entry focused mostly on signaling through Gα_i and did not show that Gα_i-independent pathways were similarly dispensable. A recent study reported that HIV-1 Env-mediated Rac-1 activation and subsequent reorganization of the actin filament network are required to facilitate membrane fusion (48). Another study reported a block of entry and postentry events with the B oligomer of PTX, which modulates signaling independent of Gα_i. Additional studies reported that the tyrosine kinase inhibitor genistein blocks HIV-1 infection in primary human macrophages and that RhoA and filamin A are required for the actin-dependent clustering of CD4 and coreceptor required for membrane fusion (1, 20, 31, 54). These findings indicate that HIV-1 Env-induced signaling plays a crucial role in Env-mediated membrane fusion.

In the current study, we elucidate that the signal transduction pathway of Env-dependent Rac-1 activation and membrane fusion is Gα_q, PLCβ, PKC, Pyk2, and Ras dependent (Fig. 1). Env-mediated Rac-1 activation and membrane fusion were attenuated in cells expressing siRNA targeted to Gα_q and Pyk2 and in cells treated with the PLC inhibitor U73122 but not in cells treated with its inactive analog U73343 or in cells expressing siRNA targeted to Gα_i and Gα_s. Rac-1 activation and fusion were also blocked by multiple PKC inhibitors, by inhibitors of intracellular Ca²⁺ release, and by the Ras inhibitor FTS. The observation that siRNA and inhibitors targeting the Gα_q pathway were inhibitory but that inhibitors and siRNA targeting the Gα_i or the Gα_s pathway were not indicates that this pathway is both necessary and sufficient to mediate membrane fusion. Using a constitutively active Rac mutant, RacV12, we have shown that Gα_q, PLCβ, PKC, and Pyk2 are necessary upstream of Rac-1 whereas Ras activation and intracellular Ca²⁺ release are required for cell-cell fusion both upstream and downstream of Rac-1 activation. Importantly, the inhibitor concentrations employed in our experiments did not alter levels of surface expression or localization of CD4 and CCR5 and also did not show any nonspecific effects on vaccinia virus infection and T7 polymerase or on other signaling pathways. Inhibitors to PKC, intracellular Ca²⁺ release, and Ras suppressed Env-dependent and virus-dependent fusion as well as HIV_YU2 infection of TZM-BL cells to comparable levels. On the other hand, VSV-G HIV-1 or A-MLV HIV-1 infection of TZM-BL cells was not blocked by the PKC, intracellular Ca²⁺ release, or Ras inhibitors, confirming that the block is specific to HIV-1 Env-mediated membrane fusion. The PLC inhibitor U73122 had a greater effect on HIV_YU2 infection of TZM-BL cells than on Env-dependent and virus-dependent fusion and also decreased VSV-G HIV-1 and A-MLV HIV-1 infection. Taken together, these data suggest that, while all of the inhibitors suppress fusion and therefore infection, U73122 can also suppress postfusion steps in the virus life cycle.

It has been shown previously that binding of CCR5 by HIV-1 gp120 activates several signaling pathways, including those involving PKC activation, Ca²⁺ elevation, Pyk2 phosphorylation, and Rac-1 activation (48-50). In the present study, we show that Env interaction with CCR5 also results in Ras activation. Env-induced Ras activation is mediated specifically by CCR5 and not by CD4 alone, because no increase in RasGTP was observed in U87.CD4.CCR5 cells stimulated with X4 Env or TAK-779-treated cells stimulated with R5 Env. Env-dependent Ras activation provides a link between CCR5 stimulation and Rac-1 activation. We have shown previously that Env-mediated Rac-1 activation is dependent on a Rac-specific GEF, most likely Tiam-1 (46), and Ras has been shown to mediate Rac-1 activation via a direct interaction with Tiam-1 or via phosphatidylinositol 3-OH kinase-mediated activation of Tiam-1 (35). The Env-mediated activation of Ras upstream of Rac-1 indicates that Tiam-1 is the RacGEF responsible for Env-mediated Rac-1 activation. RacGEFs and their interacting proteins play an important role in the selection of downstream effectors of Rac-1, leading to regulation of different cellular processes (16, 52, 60). Therefore, discovery of the RacGEF and of downstream targets of Rac-1 required for fusion will provide information on the mechanism of fusion and additional targets for therapeutic intervention.

In summary, we provide evidence that siRNA and inhibitors targeted to Gα_q and its downstream effectors prevent HIV-1 infection of TZM-BL cells as well as Env-dependent membrane fusion in U87.CD4.CCR5 cells and PBLs and virus-dependent membrane fusion of U87.CD4.CCR5 cells. HIV-1 Env interaction with CCR5 stimulates a signaling cascade involving Gα_q, PLC, PKC, intracellular Ca²⁺ release, Pyk2, and Ras that allows activation of Rac-1 and subsequent actin cytoskeleton rearrangements necessary for fusion. These results confirm and extend the implication of Rac-1 in HIV-1 infection and point to new potential target molecules of HIV-1-inhibitory drugs. Multiple PKC inhibitors are in clinical trials or are approved for clinical use to treat cancer and diabetes, including Go-6976 and Bis I used in this study (38). Multiple studies of mice with the Ras inhibitor FTS have given promising results for the treatment of various types of cancer, neurofibromatosis, and kidney disease (7, 11, 33, 34). In addition, FTS is currently in phase I clinical trials for hematologic malignancies (7). The ability of these inhibitors to specifically downregulate their target molecules without adverse side effects suggests that these inhibitors might be appropriate drugs for the treatment of HIV-1 and other infectious microbes that manipulate this pathway. This strategy of using inhibitors that disable host signaling proteins essential for pathogen survival may have a general efficacy in developing drugs to combat pathogens that acquire drug resistance.