Human Immunodeficiency Virus (HIV)-1 Viral Protein R Suppresses Transcriptional Activity of Peroxisome Proliferator-Activated Receptor γ and Inhibits Adipocyte Differentiation: Implications for HIV-Associated Lipodystrophy

Abstract

HIV-1-infected patients may develop lipodystrophy and insulin resistance. We investigated the effect of the HIV-1 accessory protein viral protein R (Vpr) on the activity of the peroxisome proliferator-activating receptor-γ (PPARγ), a key regulator of adipocyte differentiation and tissue insulin sensitivity. We studied expression of PPARγ-responsive reporter genes in 3T3-L1 mouse adipocytes. We investigated Vpr interaction with the PPAR/retinoid X receptor (RXR)-binding site of the c-Cbl-associating protein (CAP) gene using the chromatin immunoprecipitation assay as well as the interaction of Vpr and PPARγ using coimmunoprecipitation. Finally, we studied the ability of exogenous Vpr protein to enter cultured adipocytes and retard differentiation. We found that Vpr suppressed PPARγ-induced transactivation in both undifferentiated and differentiated 3T3-L1 cells. Transcriptional suppression by Vpr required an intact LXXLL coactivator motif. Vpr suppressed mRNA expression of PPARγ-responsive genes in undifferentiated 3T3-L1 cells and associated with the PPAR/RXR-binding site located in the promoter region of the CAP gene. Vpr interacted with the ligand-binding domain of PPARγ in an agonist-dependent fashion in vitro. Vpr delivered either by an expression plasmid or as protein added to media suppressed PPARγ agonist-induced adipocyte differentiation, assessed as lipid accumulation and mRNA expression of the adipocyte differentiation marker adipocyte P2 in 3T3-L1 cells. In conclusion, circulating Vpr or, alternatively, Vpr produced as a consequence of direct infection of adipocytes could suppress in vivo differentiation of preadipocytes by acting as a corepressor of PPARγ-mediated gene transcription. Vpr may alter sensitivity to insulin and thereby contribute to the development of lipodystrophy and insulin resistance observed in HIV-1-infected patients.

LIPODYSTROPHY AND METABOLIC abnormalities have emerged as a common feature of HIV-1 disease, first reported in 1998 (1) and now believed to affect approximately 50% of patients who have received long-term antiretroviral treatment (2). We have set out to explore a possible role for the HIV-1 accessory protein in the pathogenesis of this syndrome. Lipodystrophy includes a decrease in peripheral fat (face and extremity sc fat) and an increase in central fat (dorsocervical and axial fat pads and abdominal visceral fat); these changes may not occur in the same patients at the same time (3). Metabolic features include elevated serum concentrations of low-density lipoprotein cholesterol and triglycerides and decreased levels of high-density lipoprotein cholesterol, carbohydrate intolerance, insulin resistance, hyperinsulinemia, and diabetes mellitus.

The mechanisms responsible for these abnormalities are not well understood. HIV-1 protease inhibitor therapy has been implicated, possibly via inhibition of glucose transporter-4-mediated glucose transport, defective insulin signaling, and activation of lipolysis (4). Nonetheless, the syndrome also occurs in patients receiving antiviral regimens lacking protease inhibitors (5), perhaps due to mitochondrial DNA damage induced by nucleoside reverse transcriptase inhibitor therapy and consequent mitochondrial changes (6). Also, lipodystrophy and metabolic abnormalities can occur in therapy-naive patients, in whom there is an association with viral burden, at least in men (7). This suggests the possibility that a factor intrinsic to the virus might also contribute.

Adipocyte function, lipid metabolism, and peripheral tissue insulin sensitivity are critically influenced by the peroxisome proliferator-activated receptor-γ (PPARγ) (8). In the nucleus, ligand-bound PPARγ heterodimerizes with members of the retinoic X receptor (RXR) family, binds to peroxisomal proliferator response elements (PPRE) located in gene promoters, stimulates their transcription, and thus regulates cell differentiation, fat storage, and insulin sensitivity of fat and muscle. Prostaglandin J2 and linoleic acid function as endogenous PPARγ ligands with relatively low affinity. Thiazolidinediones are pharmacological PPARγ ligands and have been used extensively in the treatment of type 2 diabetes and other disorders characterized by insulin resistance.

HIV-1 encodes a 96-amino-acid accessory protein, viral protein R (Vpr), which is packaged in significant quantities into viral particles and is imported into the nucleus early after cell infection (9, 10). Vpr has at least four distinct functions. Vpr participates in the nuclear translocation of the HIV-1 preintegration complex (9, 10). Vpr arrests cells at the G2/M phase of the cell cycle, which may facilitate viral replication (11–13). Vpr induces apoptosis, at least in part by altering mitochondrial membrane potential (14). Finally, Vpr is a transcriptional activator of viral and cellular promoters (15). Vpr enhances the transcriptional activity of several steroid hormone receptors, including the glucocorticoid receptor (GR) and progesterone receptor, acting as a potent coactivator (16–18), in contrast to the adenovirus E1A, which behaves as a corepressor for several nuclear receptors including the GR (18). Vpr shares the LXXLL motif, located at amino acids 64–68, with steroid receptor coactivators, such as p160-type proteins and p300/cAMP response element binding protein-binding protein (CBP). The LXXLL motif accounts for the ability of Vpr to bind ligand-activated GR (16, 19). Vpr also interacts with the endogenous coactivators such as p300/CBP and efficiently attracts these proteins to the promoter region of glucocorticoid-responsive genes (20). Vpr circulates in HIV-1-infected individuals, readily penetrates plasma membranes of cultured cells, and may exert bystander effects in uninfected cells (21–23). Importantly, Vpr may be produced by chronically infected cells even in the presence of effective antiviral therapy (24).

We explored the hypothesis that Vpr might contribute to lipodystrophy and metabolic dysregulation by influencing the PPARγ receptor signaling system. Our results suggest that Vpr acts as a corepressor of PPARγ activity, potentially contributing to these complications.

Results

We selected mouse preadipocyte 3T3-L1 cells as a model system to examine the effect of Vpr on PPARγ activity, because these cells differentiate in response to stimuli including PPARγ ligands (25). In undifferentiated 3T3-L1 cells, Vpr suppressed ciglitazone-stimulated PPARγ activity and 15d-PGJ₂-stimulated PPARγ activity on a PPRE-containing promoter in a dose-dependent fashion (Fig. 1, A and B). Similarly, Vpr suppressed ciglitazone-stimulated PPARγ activity in differentiated 3T3-L1 (Fig. 1C) and in HeLa human cervical carcinoma cells (Fig. 1D).

Fig. 1.

Vpr Suppressed PPARγ-Induced Transcription in Undifferentiated (A and B) and Differentiated (C) 3T3-L1 and HeLa (D) Cells Cells were transfected with Vpr, PPARγ, and RXRγ expression plasmids, together with the reporter gene PPRE-TK-Luc and the normalizing gene pCMV-β-Gal.Bars represent mean ± sem values of the luciferase activity normalized for β-galactosidase activity in the absence or presence of ciglitazone (A, C, and D) or 15d-PGJ₂ (B). Vpr suppressed PPRE-mediated gene expression in a dose-dependent fashion in the presence of the PPARγ ligands ciglitazone and 15-dPDJ, in both murine adipocytes (3T3-L1 cells) and human cervical carcinoma HeLa cells. *, P < 0.01 compared with baseline (0 in the presence of ciglitazone).

We next tested the effects of Vpr on the transcriptional activity of the other PPAR family members, PPARα and -δ, and another nuclear receptor, pregnane X receptor (PXR). These two PPARs also play role in lipid and glucose metabolism, whereas PXR is essential for xenobiotic metabolism by binding to numerous bioactive compounds (26). Vpr was inactive on Wy14643-stimulated transcriptional activity of PPARα, whereas it markedly enhanced GW501516-activated PPARδ-induced transcriptional activity on a PPRE-containing promoter in undifferentiated 3T3-L1 cells (Fig. 2, A and B). Vpr suppressed rifampicin-stimulated transcriptional activity of PXR similarly to its effect on PPARγ (Fig. 2, C and D). Vpr was inactive on the transcriptional activity of the Rous sarcoma virus promoter, which does not contain binding sites for these nuclear receptors. These results indicate that Vpr has differential effects on other members of the PPAR family as well as on PXR.

Fig. 2.

Vpr Has Different Transcriptional Effects on PPARγ, PPARα, PPARδ, and PXR-Mediated Transcription in Undifferentiated 3T3-L1 Cells Cells were transfected with Vpr- and RXRγ-expressing plasmids, PPRE-TK-Luc, and pCMV-β-Gal together with PPARα-expressing (A), PPARδ-expressing (B) or PXR-expressing (C) plasmids. The Vpr-expressing plasmid RSV-Luc and pCMV-β-Gal were transfected into the cells (D). Bars represent mean ± sem values of the luciferase activity normalized for β-galactosidase activity in the absence or presence of Wy14643 (A), GW501516 (B), ciglitazone (C), or rifampicin (D). *, P < 0.01; n.s., not significant.

Given the major role of PPARγ on insulin action and adipose distribution, we focused on the effect of Vpr on the transcriptional activity of PPARγ. Because the above results were obtained in the presence of RXRγ, which heterodimerizes with PPARγ, and to isolate the effect of Vpr on PPARγ itself, we used a mammalian one-hybrid system, in which a chimeric molecule, PPARγ-Gal4 DNA-binding domain (DBD), interacts with the Gal4 response element-driven promoter. Vpr suppressed ciglitazone-stimulated GAL4-PPARγ-induced transcriptional activity in a dose-dependent fashion (Fig. 3). Vpr had no effect on this promoter in the absence of GAL4-PPARγ (data not shown), indicating that the Vpr transcriptional effect is specific for GAL4-PPARγ.

Fig. 3.

Vpr Suppressed PPARγ-Induced Transactivation in a GAL4-DBD-Mediated Mammalian One-Hybrid System in Undifferentiated 3T3-L1 Cells Cells were transfected with a plasmid that expresses PPARγ fused to a GAL4-DBD, together with a response plasmid pUAS-TK-Luc and the normalizing plasmid pCMV-β-Gal. Data represent mean ± sem of luciferase activity normalized for β-galactosidase activity. There was a dose-dependent increase in gene expression, which suggests that Vpr interacts with PPARγ, and this interaction is facilitated by the PPARγ ligand ciglitazone. *, P < 0.01 comparing the presence and absence of Vpr.

We next identified Vpr domains required for repression of PPARγ transcriptional activity. Vpr L64,67,68A had no effect on ciglitazone-stimulated PPARγ transcriptional activity, whereas VprR80A maintained suppressive activity (Fig. 4A). These results suggest that the LXXLL coactivator motif is necessary for Vpr to suppress PPARγ-induced transcriptional activity, whereas the C-terminal motif that mediates cell cycle arrest activity is dispensable.

Fig. 4.

Vpr Suppressed PPARγ-Induced Transactivation via Its LXXLL Motif Independently of Its Enhancement of GR Transcriptional Activity A, Vpr suppressed PPARγ-induced transactivation through an LXXLL motif. Undifferentiated 3T3-L1 cells were cotransfected with indicated amounts of PPARγ, RXRγ, PPRE-TK-Luc, and pCMV-β-Gal and with mutant and Vpr-expression plasmids. Bars represent values of the luciferase activity normalized for β-galactosidase activity in the absence or presence of ciglitazone. Mutant Vpr R80A, deficient in cell cycle-arrest activity, was able to suppress PPARγ-induced transactivation in undifferentiated 3T3-L1 cells. In contrast, triple-mutant Vpr lacking the corepressor motif was unable to suppress PPARγ-induced transactivation. *, P < 0.01, compared with baseline (vector in the presence of ciglitazone). B and C, Vpr suppressed PPARγ-induced transactivation independently of its coactivation of glucocorticoid-mediated gene transcription. Cells were cotransfected with PPARγ, RXRγ-pCMV-β-Gal, and MMTV-Luc expression plasmids (B) or PPRE-TK-Luc (C) in the presence or absence of Vpr-expression plasmid. Cells were subsequently treated with dexamethasone and/or RU 486 (B) and ciglitazone (C). Bars represent mean ± sem. Values of the luciferase activity were normalized for β-galactosidase. *, P < 0.01; n.s., not significant compared with the baseline (vector).

Glucocorticoids have a supportive role in adipocyte differentiation (27), and Vpr can enhance glucocorticoid action. We investigated whether Vpr potentiation of GR transcriptional activity contributes to PPARγ suppression. RU 486, a competitive inhibitor of GR that lacks activity on PPARγ, did not affect Vpr-induced suppression of PPARγ transactivation, whereas it completely suppressed GR-induced transactivation on the glucocorticoid-responsive mouse mammary tumor virus (MMTV) promoter (Fig. 4, B and C).

We investigated the effect of Vpr on the expression of endogenous PPARγ-responsive c-Cbl binding protein (CAP) gene in undifferentiated 3T3-L1 cells. CAP is predominantly expressed in insulin-sensitive tissues and positively regulates insulin action, directly associating with both the insulin receptor and the c-Cbl protooncogene product (28) Thiazolidinediones stimulate CAP expression, and thiazolidinedione-induced CAP expression correlates well with increased insulin sensitivity both in vitro and in vivo (28). We transfected 3T3-L1 cells with wild-type Vpr or Vpr L64,67,68A together with PPARγ and RXRγ and treated the cells with ciglitazone or vehicle in the absence or presence of the PPARγ antagonist GW9662 for 24 h. Ciglitazone (100 μm) stimulated CAP mRNA expression 3.5-fold. Wild-type Vpr suppressed ciglitazone-stimulated CAP mRNA expression, whereas Vpr L64,67,68A had no effect. The addition of the PPARγ antagonist GW9662 (100 nm) significantly suppressed ciglitazone-induced expression of CAP mRNA and reduced the observed repression by wild-type Vpr (Fig. 5A).

Fig. 5.

Vpr Suppressed mRNA Expression of the Endogenous PPARγ-Responsive CAP Gene in Undifferentiated 3T3-L1 Cells A, Wild-type Vpr, but not the L64,67,68A mutant, suppressed the CAP mRNA expression in undifferentiated 3T3-L1 cells. Cells were transfected with Vpr and PPARγ- and RXRγ-expression plasmids and treated with ciglitazone and/or the PPARγ antagonist GW9662 for 24 h. Total RNA was purified from the cells, and the expression of CAP and GAPDH mRNAs was determined by real-time PCR. Bars show mean ± sem of the fold induction of CAP normalized for GAPDH, compared with baseline (vector in the presence of ciglitazone and in the absence of GW9662). Ciglitazone stimulation of CAP expression was blunted by wild-type Vpr and in the presence of GW9662 but not by triple-mutant Vpr lacking the corepressor motif. B, Vpr expressed by lentivirus infection strongly suppressed ciglitazone-induced CAP mRNA expression in undifferentiated 3T3-L1 cells. Cells were infected with the lentiviruses that express Vpr or EGFP and were treated with ciglitazone and/or GW9662 for 24 h. Total RNA was purified from the cells, and the expression of CAP, PPARγ, and GAPDH mRNAs was determined by real-time PCR. Bars show mean ± sem of the fold induction of CAP or PPARγ normalized for GAPDH, compared with baseline (EGFP in the presence of ciglitazone and in the absence of GW9662). C, Vpr was expressed in cells infected with the lentivirus vector for Vpr. Cells were infected with lentiviruses expressing Vpr or EGFP, and cell lysates were run on a 4–20% SDS-PAGE gel. Vpr peptide (10 ng) was run as a control. After blotting on a nitrocellulose membrane, Vpr was visualized with the anti-Vpr antibody. D, Vpr expressed by lentivirus infection suppressed the expression of the ciglitazone-induced PPARγ-responsive aP2, PAGP, and CD36 mRNAs in undifferentiated 3T3-L1 cells. The same samples used in Fig. 5B were examined. aP2, PGAR, CD36, and GAPDH mRNAs were determined by real-time PCR. Bars show mean ± sem of the fold induction of aP2, PGAR, or CD36 mRNAs normalized for GAPDH, compared with baseline (EGFP) in the presence of ciglitazone.

We further employed the lentivirus-mediated gene delivery method for expressing Vpr and examined its effect on the endogenous CAP mRNA expression in undifferentiated 3T3-L1 cells by treating them with ciglitazone and/or GW9662 for 24 h. Because transient transfection may involve high-level expression of transgene and may disrupt cell function, we carried out similar experiments using lentiviral delivery of Vpr. Expression of Vpr, but not enhanced green fluorescent protein (EGFP), from the lentivirus strongly suppressed ciglitazone-stimulated mRNA expression of the endogenous CAP gene, whereas GW9662 potently attenuated these effects (Fig. 5B, left). Neither the lentivirus infection nor addition of ciglitazone and/or GW9662 influenced the mRNA expression of PPARγ (Fig. 5B, right). Also, the infection with the lentivirus was associated with expression of immunoreactive Vpr in these cells (Fig. 5C). In addition to the CAP gene, lentivirus-mediated Vpr expression suppressed mRNA expression of the PPARγ-responsive PPARγ angiopoietin-related (PGAR), adipocyte P2 (aP2), and CD36 genes (29–32) (Fig. 5D). Taken together, these results obtained with transfection-based and lentivirus-mediated gene delivery methods demonstrate that Vpr suppresses the transcriptional activity of PPARγ on an endogenous PPRE-containing gene in a PPARγ activation-dependent fashion.

To further examine mechanisms of Vpr-induced suppression of PPARγ-induced transcription, we tested interaction of Vpr and PPARγ using an in vitro glutathione S-transferase (GST) pull-down assay (Fig. 6A). ³⁵S-labeled Vpr interacted with bacterially produced full-length PPARγ and with PPARγ ligand-binding domain (LBD) in a ciglitazone-dependent fashion. In contrast, Vpr did not bind GST alone in the absence of PPARγ. Because Vpr is known to bind directly to GR via the LXXLL motif located between amino acids 64 and 68, we examined binding of LXXLL motif-defective VprL64,67,68A to GST-fused PPAR and found that this mutant did not bind either full-length PPARγ or PPARγ LBD. These results indicate that Vpr interacts directly with PPARγ LBD through its LXXLL motif in a ligand-dependent fashion.

Fig. 6.

Vpr Interacted with PPARγ in Vitro and in Vivo Vpr attenuates attraction of p300 coactivator to the PPARγ-responsive promoter, whereas it does not have autonomous transactivation or transrepression activity in undifferentiated 3T3-L1 cells. A, Wild-type Vpr, but not LXXLL defective Vpr mutant, bound the LBD of PPARγ in a ligand-dependent fashion in a GST pull-down assay. In vitro translated and labeled wild-type Vpr or VprL64,67,68A was incubated with bacterially produced GST-fused full-length PPARγ or PPARγ LBD immobilized on GST beads. Results of the GST pull-down assay are shown in the top panels, whereas expression of the GST-fusion proteins is shown in the bottom panels. B, Wild-type Vpr, but not L64,67,68A mutant, was attracted to the PPRE/RXR-binding site of the CAP promoter in a ciglitazone-dependent fashion in undifferentiated 3T3-L1 cells. Cells were transfected with wild-type Vpr, VprL64,67,68A, and PPARγ- and/or RXRγ-expression plasmids and treated with ciglitazone or vehicle. Twenty-four hours after addition of ciglitazone, the cells were fixed, and the ChIP reaction was performed with anti-Vpr, anti-PPARγ, or control antibody. The portion of the CAP promoter that contains one PPAR/RXR-binding site was amplified by PCR. Results obtained in the SYBR-Green real-time PCR for quantitatively evaluating the ChIP results are shown in the right panels. Bars represent mean ± se values of fold precipitation of the CAP PPRE/RXR-binding site, compared with the baseline. *, P < 0.01; n.s., not significant, compared results to baseline (vector in the presence of ciglitazone). C, Wild-type Vpr attenuated ciglitazone-induced association of p300 to the CAP promoter in undifferentiated 3T3-L1 cells. Cells were transfected with Vpr and PPARγ- and RXRγ-expression plasmids and treated with ciglitazone or vehicle. Twenty-four hours after addition of ciglitazone, the cells were fixed, and the ChIP reaction was performed with anti-Vpr, anti-PPARγ, anti-p300, or control antibody. The portion of the CAP promoter that contains one PPAR/RXR-binding site was amplified by PCR with a specific primer pair, and obtained gel images on 3% DNA gels are shown in the left panel. Results obtained in the SYBR-Green real-time PCR for quantitatively evaluating the ChIP results are shown in the right panels. Bars represent mean ± se values of fold precipitation of the CAP PPRE/RXR-binding site, compared with the baseline. *, P < 0.01; n.s., not significant, compared with the baseline (vector in the presence of ciglitazone). D, Vpr does not have intrinsic transactivation or transrepression activity in undifferentiated 3T3-L1 cells. Undifferentiated 3T3-L1 cells were transfected with pM, pM-Vpr, or pGLA4-Vpr16, together with GAL4-responsive p17mer-tk-Luc and CMV-β-galactosidase. Bars show the means ± sem. *, P < 0.01; n.s., not significant, compared with the baseline (GAL4).

Because the LXXLL coactivator motif is required for Vpr to suppress PPARγ-mediated transactivation and binding to PPARγ in vitro, we used the chromatin immunoprecipitation (ChIP) assay to examine whether Vpr is attracted to the promoter region of the CAP gene via PPARγ. Undifferentiated 3T3-L1 cells were transfected with wild-type Vpr or Vpr L64,67,68A, together with PPARγ and RXRγ, and were treated with ciglitazone (100 μm). As a positive control, we demonstrated that nuclear extract subjected to immunoprecipitation with monoclonal anti-PPARγ antibody showed amplification of the CAP promoter region that contains PPAR/RXR binding sites, indicating that the PPARγ occupied the PPRE site both in the absence and presence of ciglitazone (Fig. 6B). Nuclear extracts from cells transfected with wild-type Vpr, but not cells transfected with Vpr L64,67,68A, supported amplification of the CAP PPAR/RXR binding site in a ciglitazone-dependent fashion. These results indicate that Vpr is attracted in vivo to the PPAR/RXR-binding site of the endogenous CAP gene via its LXXLL motif, likely after binding to PPARγ.

Previously, it has been reported that Vpr enhances GR-induced transcriptional activity by facilitating attraction of p300/CBP to the GR-bound MMTV promoter via direct interaction (25). To further examine the mechanism of Vpr-induced suppression of PPARγ transcriptional activity, we examined association of p300 on the CAP PPAR/RXR binding site (Fig. 6C). In the absence of Vpr, precipitation of the nuclear extract with anti-p300 antibody supported amplification of the PPAR/RXR binding site in a ciglitazone-dependent fashion. In contrast, amplification of the PPAR/RXR binding site was greatly attenuated in the presence of Vpr. These results indicate that Vpr inhibits attraction of p300 to the CAP PPAR/RXR binding site induced by agonist-activated PPARγ.

We investigated possible intrinsic transcriptional activity of Vpr, independent of its effects of PPARγ, by directly tethering Vpr on the promoter via GAL4 DBD (Fig. 6D). GAL4-Vpr showed no transcriptional activity on the GAL4-responsive promoter, similar to GAL4-DBD alone, whereas a positive control GAL4-DBD-VP16 activation domain fusion strongly activated the transcription. These results indicate that Vpr functions as an adaptor or negative scaffold for suppressing PPARγ-induced transcriptional activity.

To examine the effect of Vpr on a PPARγ-dependent cell phenotype, we tested whether Vpr could suppress ciglitazone-dependent lipid uptake by differentiated 3T3-L1 cells. Vpr was expressed intracellularly from the plasmid pIRES2-EGFP-Vpr, which expresses Vpr and EGFP separately under the control of the same promoter. pIRES2-EGFP was transfected as negative control. In the absence of ciglitazone, 3T3-L1 cells exhibited a fibroblast phenotype, with an elongated shape and scant cytoplasmic lipid. In the presence of ciglitazone, these cells differentiate into adipocytoid cells, with round shape and abundant lipid droplets. Despite treatment with ciglitazone, Vpr-transfected cells, identified as expressing EGFP, maintained a fusiform shape and did not accumulate lipid (Fig. 7). Nontransfected cells in the same culture differentiated into adipocytoid cells. Thus, Vpr antagonizes the action of PPARγ in vivo, suppressing adipocytic differentiation of 3T3-L1 cells induced by ciglitazone.

Fig. 7.

Vpr Delivered by Transfection Suppresses Ciglitazone-Induced Lipid Accumulation and Differentiation of 3T3-L1 Cells Undifferentiated 3T3-L1 cells were transfected with pIRES2-EGFP-Vpr or the control vector pIRES2-EGFP and treated with 10 μm ciglitazone or vehicle for 48 h as indicated. Nomarski image and fluorescence signals from EGFP and lipid stained with Nile red were obtained. Ciglitazone enhanced cellular lipid uptake, whereas Vpr-expressing cells failed to accumulate lipid in response to ciglitazone. Arrows in the two panels are directed at cells that have taken up the Vpr-expressing plasmid (presence of green fluorescence) but are lacking the differentiated adipocyte phenotype (absence of lipid accumulation).

Vpr might be delivered to adipocytes in at least two ways. First, it has been recently suggested that adipocytes express CD4, CXCR4, and CCR5 and may be susceptible to HIV-1 infection (33). Second, Vpr may be delivered from the circulation. Vpr is detected in the sera of HIV-1-infected patients and may enter cells by a basic peptide-mediated, receptor-independent mechanism (21–23, 34). Therefore, we directly tested the ability of 3T3-L1 cells to accumulate Vpr from the surrounding medium. Addition of fluorescence-labeled Vpr, but not the control protein p6 Gag, increased cellular fluorescence in a dose-dependent fashion (Fig. 8A). The subcellular distribution of Vpr was examined, and evidence of translocation of Vpr was provided by Western blot analysis. Immunoreactive Vpr was detected in the nuclear fraction of the cells treated and untreated with ciglitazone (Fig. 8B). This result suggests that extracellularly administered synthetic Vpr peptide was taken up by the cells and translocated to the nucleus.

Fig. 8.

Extracellularly Administered Vpr Suppressed PPARγ-Induced Lipid Accumulation in 3T3-L1 Cells A, Undifferentiated 3T3-L1 cells took up extracellularly administered fluorescence-labeled Vpr peptide in a dose-dependent fashion. 3T3-L1 cells were incubated with indicated concentrations of fluorescein isothiocyanate-labeled Vpr or p6Gag and cells with the fluorescein isothiocyanate-labeled Vpr were sorted by fluorescence-activated cell sorting. B, Western blot analysis detected Vpr in the nucleus of undifferentiated 3T3-L1 cells exposed to the synthetic Vpr peptide. Undifferentiated 3T3-L1 cells were incubated with 100 ng/ml Vpr and/or 10 μm ciglitazone for 5 d, and nuclear and cytoplasmic fractions were analyzed by Western blot using antibodies against Vpr and histone 1 (the latter to confirm nuclear fractionation). C, Extracellularly administered Vpr inhibits ciglitazone-induced lipid accumulation in 3T3-L1 cells. 3T3-L1 cells were cultured with 100 ng/ml synthetic Vpr peptide and/or 10 μm ciglitazone for 5 d. The cells were analyzed for lipid accumulation with Oil Red O staining. The left panel indicates representative images of Oil Red O staining, whereas the right panel demonstrated accumulation of Oil Red O-stained area. Bars show mean ± sem of the fold induction of the Oil Red O-stained area. *, P < 0.05. Magnification, ×10. D, Extracellularly administered synthetic Vpr peptide suppresses ciglitazone-induced aP2 mRNA expression in undifferentiated (left panel) and differentiated (right panel) 3T3-L1 cells. Undifferentiated and differentiated 3T3-L1 cells were incubated with 100 ng/ml synthetic Vpr peptide and/or 10 μm ciglitazone for 5 d. Total RNA was purified from the cells, and mRNA expression of aP2 and RPLP0 were determined by real-time PCR. Cell differentiation by ciglitazone resulted in a higher concentration of aP2 mRNA as compared with undifferentiated cells treated with ciglitazone. Bars show mean ± sem of the fold induction of aP2 mRNA expression normalized for those of RPLP0. *, P < 0.01; n.s., not significant, by comparison between results obtained in the absence and presence of ciglitazone.

Finally, we examined the effect of extracellular Vpr on adipocyte phenotype. First, we added synthetic Vpr peptide into media of cultured 3T3-L1 cells and examined accumulation of lipid as monitored by Oil Red O staining. Ciglitazone significantly increased lipid accumulation, and Vpr tended to reduce this toward control levels (Fig. 8C). Next, we investigated the effect of extracellular Vpr on aP2 mRNA expression in undifferentiated and differentiated cells. PPARγ regulates aP2 expression through a PPARγ/RXR response element located in their promoter region A (30, 31). Incubation with synthetic Vpr peptide significantly suppressed ciglitazone-induced aP2 mRNA expression in both undifferentiated (Fig. 8D, right) and differentiated (Fig. 8D, left) 3T3-L1 cells. By comparison, Vpr had minimal effect on aP2 expression induced by dexamethasone/isobutyl-1-methylxanthine/insulin medium (data not shown). These results indicated that soluble Vpr suppresses PPARγ-mediated gene expression.

Discussion

Here we have demonstrated that HIV-1 Vpr suppressed ligand-stimulated PPARγ activity on a PPARγ-responsive promoter in a dose-dependent fashion in undifferentiated and differentiated 3T3-L1 adipocytes and in HeLa cells. This effect of Vpr was also observed in a mammalian one-hybrid system in which PPARγ was tethered to the GAL4-responsive element-driven synthetic promoter through GAL4-DBD. Furthermore, Vpr suppressed PPARγ-responsive CAP mRNA expression. The ChIP assay indicated that Vpr associated with the CAP promoter region in a ciglitazone-dependent fashion, possibly through direct binding to PPARγ LBD. Vpr suppressed attraction of p300 to the CAP promoter that was observed by addition of ciglitazone in the absence of Vpr. Extracellular Vpr entered cell nuclei and suppressed lipid accumulation and aP2 mRNA expression.

These results suggest the possibility that adipocytes exposed to Vpr as a consequence of cellular infection with HIV-1 or as a consequence of uptake of Vpr from plasma may be adversely affected by Vpr-mediated suppression of PPARγ-mediated gene transcription (35–39). In preliminary work using mass spectrometry, we have detected Vpr peptide in adipose and liver tissue obtained at autopsy from two HIV-seropositive patients, whereas Vpr peptides were not detected in autopsy from one control HIV-seronegative individual (Phillips, T., and J. Kopp, unpublished data).

The mechanisms of HIV-1-associated lipodystrophy are not well understood but appear multifactorial. Both HIV-1 infection and the use of protease inhibitors have been suggested to contribute to the pathogenesis (40, 41). Our results suggest that Vpr may contribute to the lipodystrophic changes seen in these patients, by inhibiting differentiation and lipid accumulation as a consequence of PPARγ signaling pathway suppression. Vpr may also contribute to insulin resistance, because activation of PPARγ is correlated with increased insulin sensitivity in muscle and liver (42, 43), whereas suppression of CAP expression could contribute to insulin resistance.

Patients with HIV-1 lipodystrophy and metabolic syndrome share many features with patients carrying PPARγ inactivating mutations (44). Common features include lipoatrophy of the extremities and lipohypertrophy in central/visceral adipose stores, hypercholesterolemia, hypertriglyceridemia, insulin resistance, and diabetes. Although there are differences in regional fat distribution, the similarities lend support to the hypothesis that PPARγ dysfunction contributes to the pathogenesis of HIV-1 lipodystrophy and metabolic syndrome.

Vpr suppresses PPARγ transactivation through its LXXLL motif. Vpr was attracted to the endogenous PPARγ-responsive promoter in a ChIP assay and bound the PPARγ LBD in a ligand-dependent fashion in a GST pull-down assay. Vpr did not demonstrate autonomous transactivation or transrepression activity, whereas it inhibited attraction of the p300 coactivator to the PPARγ-responsive promoter. It appears that Vpr suppressed PPARγ activity by interacting directly through its LXXLL motif and thus preventing accumulation of transcriptional intermediate molecules including p300/CBP at the promoter region, perhaps by acting as a negative scaffold.

The LXXLL motif has been shown previously to be necessary for Vpr to serve as a coactivator of GR-mediated gene transcription (16). The mechanism by which Vpr serves as a coactivator in one setting (GR-mediated transcription) and a corepressor in another setting (PPARγ-mediated transcription) remains an enigma, although it is not unprecedented. Nuclear factors may function as coactivators or corepressors depending upon the particular cellular environment (16, 45). Indeed, the LXXLL motif is present in both coactivators and corepressors (46). PPARγ has several distinct characteristics compared with GR that may provide clues into the different functional roles of Vpr (47–49). First, PPARγ forms a heterodimer with other nuclear receptors, whereas the GR usually acts as a homodimer. Second, ligand-bound PPARγ stimulates gene transcription via its interaction with coactivators, whereas ligand-free PPARγ suppresses gene transcription by forming a complex with corepressors (19). These differences between GR- and PPARγ-induced transcriptional activity together with yet unknown mechanisms specific to adipocytes might underlie Vpr-induced suppression of PPARγ transcriptional activity.

In conclusion, we propose that Vpr may contribute to HIV-1-associated lipodystrophy and metabolic syndrome via antagonism of PPARγ activity. Because these disorders are not common in therapy-naive patients, it may be that Vpr generally contributes as a predisposing factor rather than as a sufficient cause.

Materials and Methods

DNA Constructs

Plasmids encoding wild-type-Vpr (pCDNA3-Vpr) and L64,67,68A and R80A Vpr mutants were described previously (16). pIRES2-EGFP-Vpr was constructed by subcloning Vpr cDNA into pIRES2-EGFP (Clontech, Mountain View, CA). pM-Vpr, which expresses wild-type Vpr fused to the GAL4 DBD, was described previously (20). pGAL4-Vpr16, which expresses the VP16 activation domain fused to the GAL4 DBD, was a gift from Dr. Y. Shi (Harvard Medical School, Boston, MA). pCMV-PPARα and -PPARδ were gifts from Dr. R. M. Evans (Salk Institute, San Diego CA). pCDNA3-hPXR and RSV-Luc were gifts from Drs. F. J. Gonzalez [National Institutes of Health (NIH), Bethesda, MD) and G. N. Pavlakis (NIH, Frederick, MD)], respectively. pCDNA3-PPARγ and pGAL4-PPARγ, which, respectively, express human PPARγ and a PPARγ fused with a GAL4 DBD, and pUAS-TK-Luc, which contains the luciferase gene under the control of two GAL4-responsive elements linked to the proximal portion of the herpes simplex thymidine kinase (HSP-TK) promoter, were gifts from Dr. V. K. K Chatterjee (Cambridge University, Cambridge, UK). pCDNA3-RXRγ was constructed by introducing the human RXRγ coding sequence from CMV27103 (provided by Dr. W. Lamph, Ligand Pharmaceutical, San Diego, CA) into EcoRI and HindIII sites of pCDNA3 (Invitrogen, Carlsbad, CA). pPPRE-TK-luc, which contains the luciferase gene under the control of synthetic PPRE linked to the proximal portion of the HSV-TK promoter, was a gift of Dr. A. D. Miller (Fred Hutchinson Cancer Center, Seattle, WA). pGEX-4T3-hPPARγ (full-length) and pGEX-4T3-hPPARγ-LBD, which contain the coding sequences of full-length or LBD (amino acids from 174–475) of human PPARγ, were constructed by subcloning the corresponding cDNA sequences into pGEX-4T3 (Amersham Pharmacia Biotechnology, Piscataway, NJ). pRShGRα and pMMTV-Luc were gifts from Dr. R. M. Evans (Salk Institute, La Jolla, CA) and Dr. G. L. Hager (National Cancer Institute, Bethesda, MD), respectively. p17mer-tk-Luc, which expresses luciferase under the control of the GAL4-responsive elements, was a gift from Dr. M.-J. Tsai (Baylor College of Medicine, Houston, TX). Plasmids pCMV-β-Gal and pM were purchased from Stratagene (La Jolla, CA) and Clontech, respectively. Ciglitazone and 15-deoxy-12,14-prostaglandin J₂ (15d-PGJ₂) were purchased from Biomol Research Laboratories (Plymouth Meeting, PA).

Cell Cultures

Murine 3T3-L1 preadipocytes (American Type Culture Collection, Rockville, MD) were maintained in DMEM (Life Technologies, Gaithersburg, MD) supplemented with 10% fetal bovine serum, 2 mm l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, and 8 mg/liter biotin. Confluent 3T3-L1cell cultures were differentiated with 1 μm dexamethasone (Sigma Chemical Co., St. Louis, MO), 0.5 mm 3-isobutyl-1-methylxanthine, (Sigma), and 5 μg/ml insulin (Life Technologies) (50).

Transient Transfection and Reporter Assays

3T3-L1 cells and HeLa cells, after culture to 50% confluence, were transfected using Lipofectamine 2000 (Life Technologies) in serum-free Opti-MEM (Invitrogen). Plasmids included 0–3 μg/well of wild-type or mutant Vpr expression plasmids; 0.2 μg/well of pCMV-PPARα, pCMV-PPARδ, pCDNA3-PPARγ, and pCDNA3-RXRγ1; and 0.1 μg/well of pPPRE-TK-Luc and pCMV-β-Gal each. In a separate control experiment, 0.2 μg/well of pGAL4-PPARγ and pCDNA3-RXRγ1 and 0.1 μg/well of pUAS-TK-Luc and CMV-β-Gal were transfected along with pCDNA3-Vpr. Also, 1 μg/well of pM, pM-Vpr, or pGAL4 was transfected with 1 μg/well of p17mer-tk-Luc and 0.5 μg/well of pSV40-β-Gal in undifferentiated 3T3-L1 cells. Six hours after transfection, medium was replaced with culture medium supplemented with fetal bovine serum and ciglitazone. After 24 h, cells were lysed in cell lysis buffer (Promega, Madison, WI), and luciferase and β-galactosidase activities were determined (16). Each experiment was performed in triplicate and repeated at least three times. Results from one representative experiment are presented.

Construction and Lentiviral Vector that Expresses Vpr or GFP

A codon-modified DNA sequence that encodes the Vpr protein from the HIV-1 pNL4-3 strain was synthesized by GeneArt (Toronto, Ontario, Canada). This Vpr cDNA was subcloned into the multi-cloning site of the lentiviral expression plasmid pTRIP-poly to produce pTRIP-EF-1-Vpr. The lentiviral vector expressing the EGFP (pTRIP-EF-1-EGFP) was also produced and used as a control for Vpr. These plasmids were then cotransfected into HEK 293T cells, together with the Vpr-deficient packaging plasmid R8.91ΔVpr and the vesicular stomatitis virus G protein (VSV-G)-expressing plasmid pMD.G to produce lentiviral particles with nonspecific host selectivity. Both of these plasmids were gifts from Dr. D. Trono (Swiss Institute of Technology, Lausanne, Switzerland).

Detection of CAP, aP2, PGAP, and CD36 mRNA by Real-Time PCR

3T3-L1 cells were transfected with pCDNA3-Vpr or pCDNA3-Vpr L64,67,68A together with pCDNA3-PPARγ and pCDNA3-RXRγ by using the Nucleofector System (Amaxa GmbH, Cologne, Germany) and protocol U-30. A total of 1 × 10⁶ 3T3-L1 cells were resuspended in solution R (Amaxa) containing 5 μg of the indicated plasmids. Transfection efficiency was approximately 80%. Lentiviruses that express Vpr or EGFP were added to the cell culture medium. Twenty-four hours after either the transfection or the infection, cells were treated with 100 μm ciglitazone and/or the PPARγ antagonist 0.1 μm GW9662 (Alesis, San Diego, CA). After an additional 24 h, total RNA was purified using Trizol (Life Technologies) and subjected to reverse transcription.

To detect mRNA levels of the mouse CAP, aP2, PGAR, CD36, PPARγ, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH), specific primer pairs were used: CAP forward 5′-CATCCATTGGAAGACCTTG-3′ and reverse 5′-CTTTGTCTGGTCCATGGTTGC-3′; aP2 forward 5′-CTCCAGTGAAAACTTCGATG-3′ and reverse 5′-CTCTGACCGGATGGTGAC-3′; PGAR forward 5′-GAGACTCTGCAGAGTTTGC-3′ and reverse 5′-CTGGCTCTGAAGATTCTG-3′; CD36 forward 5′-GTTGGAACAGAGGATGACAAC-3′ and reverse 5′-CAAGTCTTCTATGTTCCAAAC-3′; PPARγ forward 5′-CTATGGAGTTCATGCTTGTG-3′ and reverse 5′-CACTGCAAGGCATTTCTG-3′; and GAPDH forward 5′-GTGGAGTCTACTGGTGTCTTC-3′ and reverse 5′-CATCACTGCCACCCAGAAGAC-3′. Quantitative PCR was performed in an ABI Prism 5500 SDS LightCycler (Applied Biosystems, Foster City, CA) and included heat activation of the Taq polymerase (10 min at 95 C) and 60 cycles (denaturing for 15 sec at 95 C, annealing/extension for 1 min at 60 C). Reactions were carried out in triplicate using the SYBR-Green PCR Master Mix (Applied Biosystems). The dissociation curves of primer pairs showed a single peak, and amplicons had a single DNA band of the predicted size in an agarose gel analysis (data not shown). Threshold cycle values of CAP, aP2, PGAP, CD36, and PPARγ were normalized for those of GAPDH, and relative mRNA expression is presented as fold induction over the baseline.

Expression of aP2 mRNA during Adipocytic Differentiation of 3T3-L1 Cells

aP2 mRNA was measured in undifferentiated and differentiated 3T3-L1 cells. Undifferentiated 3T3-L1 cells were treated with 300 ng/ml Vpr (full-length Vpr peptide chemically synthesized) (23, 34) and/or 10 μm ciglitazone. For ciglitazone-induced differentiation, 3T3-L1 cells were cultured for 48 h and treated with 300 ng/ml Vpr and/or 10 μm ciglitazone on alternate days. Five days later, total RNA from undifferentiated and differentiated cells was purified and reverse transcribed, and mRNA levels of the mouse aP2 gene and the control ribosomal phosphoprotein 0 (RPLP0) gene were determined with real-time PCR. Conditions for aP2 amplification were identical to those described above for CAP and aP2. The primer pair used for amplifying RPLP0 was forward 5′-GAGGACCTCACTGAGATTCG-3′ and reverse 5′-CTGGAAGAAGGAGGTCTTCTC-3′.

ChIP Assay

3T3-L1 cells were transfected with pCDNA3-Vpr or pCDNA3-Vpr L64,67,68A together with pCDNA3-PPARγ and pCDNA3-RXRγ using the Nucleofector System. Twenty-four hours after transfection, cells were treated with 100 μm ciglitazone or vehicle (ethanol). After an additional 24 h, cells were processed for ChIP assay using a Chromatin Immunoprecipitation Kit (Upstate, Charlottesville, VA). Samples containing DNA/protein complexes were incubated overnight with anti-Vpr monoclonal antibody 9F12 (provided by U. Schubert and J. Yewdell, NIH, Bethesda, MD), anti-PPARγ antibody, anti-p300 antibody, or mouse control IgG (all from Santa Cruz Biotechnologies, Santa Cruz, CA). Immune complexes were collected by adding protein A slurry, and cross-linked DNA and bound proteins were uncoupled by heating at 65 C for 4 h.

The promoter region (−1150 to −1048) of the mouse CAP gene, which contains a single PPRE/RXR-binding site (located at −1099 to −1082), was amplified by Taq DNA polymerase (Applied Biosystems). PCR was carried out using the following primers: forward 5′-TGTGCCTCAGGTGACTATTC-3′ and reverse 5′-GGAGGCATTTTCTTAATTGTGGTTCC-3′. PCR involved 40 cycles as follows: denaturing, 1 min at 94 C; annealing, 1 min at 50 C; and elongation, 1 min at 72 C. For quantitative evaluation of the ChIP results, SYBR-Green real-time PCR was performed using SYBR-Green PCR Master Mix (Applied Biosystems) and a 5500 Real-Time PCR System (Applied Biosystems). Obtained threshold cycle values of ChIP samples were normalized for those of corresponding inputs, and their relative precipitation was demonstrated as fold precipitation above the baseline.

In Vitro Binding Assay

³⁵S-labeled Vpr was generated by in vitro transcription and translation reaction with wheat germ extract (Promega) and pCDNA3-Vpr as template. Vpr was tested for interaction with the GST-fused full-length or LBD of human PPARγ immobilized on glutathione-Sepharose beads in a buffer containing 50 mm Tris-HCl (pH 8.0), 50 mm NaCl, 1 mm EDTA, 0.1% Nonidet P-40, 10% glycerol, and 0.1 mg/ml BSA at 4 C for 1.5 h. After vigorous washing, proteins were eluted and separated on 14% SDS-PAGE gels. Approximately 10% total input of labeled Vpr was loaded as a control.

Lipid Accumulation in Differentiated 3T3-L1 Cells

Undifferentiated 3T3-L1 cells were transfected with 0.5 μg/well of pIRES2-EGFP-Vpr or the control plasmid pIRES2-EGFP. The former plasmid expresses Vpr and the EGFP under the control of the cytomegalovirus promoter and contains an internal ribosome entry site (IRES). Six hours after transfection, 10 μm ciglitazone was added, and the cells were cultured for 48 h. Intracellular lipid accumulation was detected by exposing cells to 0.4 μm Nile red for 30 sec. The cells were examined using Nomarski optics and epifluorescence microscopy using appropriate filters to detect Nile red and EGFP.

Lipid Accumulation in Undifferentiated 3T3-L1 Cells

Undifferentiated 3T3-L1 cells were treated with medium, 10 μm ciglitazone, 100 ng/ml Vpr, or both on d 1 and 5. On d 6, cultures were washed with PBS, fixed with buffered formalin, and stained with Oil Red O to detect neutral lipids. Photomicrographs were taken at random at ×10, and the fractional lipid area was determined using Image J software (NIH, Bethesda, MD).

Detection of Incorporation of Fluorescence-Labeled Vpr Peptide in 3T3-L1 Cells

Undifferentiated 3T3-L1 cells were resuspended at 1 × 10⁶ cells/ml in serum-containing culture medium. Cells were incubated overnight at 37 C in the presence of ALEXA-488-labeled (Molecular Probes, Eugene, OR) synthetic Vpr peptide or full-length 52-amino-acid HIV-1 Gag protein p6^gag as control. Cells were washed in PBS, fixed in 2% formaldehyde for 30 min, resuspended in PBS, and examined with a FACScalibur (Becton Dickinson, San Jose, CA). Data were analyzed using CellQuest (Becton Dickinson) and FlowJo (Tree Star, San Carlos, CA) software.

Localization of Vpr in the Nucleus of 3T3-L1 Cells

Undifferentiated 3T3-L1 cells were treated overnight with 100 ng/ml Vpr, and 24 h later 10 μm ciglitazone was added. Two days later, fresh medium with the same supplements was added. After 24 h, nuclear and cytosolic fractions were obtained using a cell fractionation kit (BioVision, Mountain View, CA). Identical amounts (60 μg) of protein were run on 12% Bis-Tris gels (Invitrogen), blotted on polyvinylidene difluoride membranes, and exposed to rabbit polyclonal anti-Vpr serum (raised against Vpr 1–46) or rabbit polyclonal anti-histone 1 IgG (Santa Cruz Biotechnologies). Blots were visualized by chemiluminescence using Super Signal West Dura substrates (Pierce Biotechnology, Rockford, IL) and autoradiography.

Statistical Analyses

Data are presented as mean ± sem. Statistical analyses were performed by one-way ANOVA and by Student’s t test as appropriate, using Prism software (GraphPad, San Diego, CA). P < 0.05 was taken as significant.

Acknowledgments

We thank Drs. V. K. K. Chatterjee, R. M. Evans, G. Hager, W. Lamph, A. D. Miller, F. J. Gonzalez, G. N. Pavlakis, D. Trono, and M.-J. Tsai for generously providing plasmids and reagents. We also thank Dr. Ashok Balasubramanyam for helpful discussion and Dr. Sonia Doi for her help with Oil Red O staining.

This work was supported by the Intramural Research Program of the National Institute of Diabetes and Digestive and Kidney Diseases, National Institute of Child Health and Human Development, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD, and Grant RO1 DK-59537.

Disclosure Statement: The authors have nothing to disclose.

¹

S.S. and T.K. have contributed equally to this manuscript.

Abbreviations

 
• aP2,

  Adipocyte P2;

 
• CAP,

  c-Cbl-associating protein;

 
• CBP,

  cAMP response element binding protein-binding protein;

 
• ChIP,

  chromatin immunoprecipitation;

 
• DBD,

  DNA-binding domain;

 
• EGFP,

  enhanced green fluorescent protein;

 
• GR,

  glucocorticoid receptor;

 
• GST,

  glutathione S-transferase;

 
• LBD,

  ligand-binding domain;

 
• MMTV,

  mouse mammary tumor virus;

 
• PGAR,

  PPARγ angiopoietin-related;

 
• PPARγ,

  peroxisome proliferator-activated receptor-γ;

 
• PPRE,

  peroxisomal proliferators response elements;

 
• PXR,

  pregnane X receptor;

 
• RXR,

  retinoid X receptor;

 
• Vpr,

  viral protein R.