Endothelial Cells Enhance Human Immunodeficiency Virus Type 1 Replication in Macrophages through a C/EBP-Dependent Mechanism

Abstract

Macrophages are early targets of human immunodeficiency virus type 1 (HIV-1) infection and serve as potential reservoirs for long-term infection. Through inflammatory mediators and direct cell contact, infected macrophages interact with neighboring cell populations, such as the endothelium, which create a microenvironment favorable for HIV-1 replication. We hypothesize that the transcriptional activator C/EBPβ is critical for macrophages to respond to endothelial cell-derived signals. We show that endothelial cells significantly enhance C/EBPβ binding activity and HIV-1 replication in macrophages. This increase in HIV-1 transcription is due to cell-cell contact as well as the production of soluble factors, mediated in part by ICAM-1 and interleukin 6, respectively. Furthermore, C/EBP factors are necessary for endothelial cell-dependent activation of HIV-1 transcription in macrophages, and HIV-1 induction can be inhibited by a C/EBP dominant-negative protein. In addition, C/EBP binding sites are necessary for efficient LTR activity and HIV-1 replication in the presence of endothelial cells. Taken together, these results indicate that endothelial cells, through the activation of C/EBPβ, provide a microenvironment that supports HIV-1 replication in monocytes/macrophages.

Macrophages are primary targets of human immunodeficiency virus type 1 (HIV-1) infection and serve as potential reservoirs for long-term infection due to their resistance to cytopathic effects of the virus (18). Furthermore, infected macrophages aberrantly express a variety of factors, including inflammatory cytokines, which directly or indirectly contribute to a number of AIDS-associated diseases, including B-cell lymphoma, Kaposi's sarcoma, interstitial pneumonitis, and dementia (9, 19, 25, 29, 32, 48, 55).

Although it is unclear whether endothelial cells are susceptible to HIV-1 infection, they clearly promote the onset of many AIDS-related pathologies (19, 36, 48). Through direct cell-cell contact between monocytes/macrophages and endothelial cells or the production of soluble factors, a microenvironment that is potentially favorable for HIV-1 replication is created. For example, endothelial cells may actively recruit HIV-1-infected cells and potential target cells, such as macrophages, to various tissue sites. Interactions between macrophages and endothelial cells would then be expected to lead to cellular activation, cytokine production, enhanced HIV-1 expression, and ultimately increased tissue damage (22, 52). However, the mechanisms by which endothelial cells affect HIV-1 expression in macrophages and establish a permissive microenvironment for HIV-1 infection and replication are not well understood.

Upon activation and differentiation of monocytes/macrophages, transcriptional activators such as NF-κB and C/EBPβ are induced (2, 3, 4, 38). These transcription factors can physically interact to synergistically regulate the expression of inflammatory cytokines and chemokines including interleukin 1β (IL-1β), IL-6, IL-8, and tumor necrosis factor alpha (TNF-α), as well as HIV-1 transcription (21, 33, 41, 42, 43, 44, 47, 50). Furthermore, inflammatory cytokines can feed back in an autostimulatory loop to induce HIV-1 transcription (15, 21, 33, 39, 41, 42, 43, 44). NF-κB has been shown to be important for HIV-1 transcription and activation of CD4⁺ T cells as well as persistent and activated viral replication in monocytes (24, 31). Although C/EBPβ is not necessary for HIV-1 transcription in CD4⁺ T cells, it is essential for HIV-1 replication in monocytes/macrophages (27, 28).

C/EBPβ is a member of the C/EBP family of transcription factors that share homology in their leucine zipper dimerization and basic DNA binding domains. Included in this family are transcriptional activators as well as dominant-negative regulators, such as liver inhibitory protein (LIP) (8, 14, 62). Functional C/EBPβ sites are required for efficient long-terminal-repeat (LTR) activity and HIV-1 replication in monocytes/macrophages (26, 27, 28, 59). In addition, establishment of HIV-1 infection and induction of proviral expression in monocytes/macrophages require endogenous C/EBP factors (27, 28). Although NF-κB has been shown to be involved in endothelial cell-induced activation of HIV-1 transcription in infected cells (20, 52), the role of C/EBPβ in macrophage-endothelial cell interactions is not known.

One potential mechanism by which endothelial cells create a microenvironment that is favorable for HIV-1 replication is by inducing transcription factors such as NF-κB or C/EBPβ in macrophages through either direct cell contact or soluble factors (20). In this study, we specifically address the role of C/EBPβ in the regulation of HIV-1 expression in response to endothelial cell-derived signals. The human promonocytic cell line U937 and primary macrophages were infected in the absence or presence of endothelial cells to investigate alterations in cytokine expression, cell function and viral replication. In these studies, the presence of endothelial cells significantly enhanced HIV-1 replication in macrophages. Most importantly, the ability of macrophages to upregulate HIV-1 in response to endothelial cell-derived signals was shown to be dependent upon the presence of C/EBP binding sites in the LTR and functional C/EBPβ protein.

MATERIALS AND METHODS

Cell lines.

U937 and U1 promonocytic cell lines and primary human umbilical vein endothelial cells (HUVEC) (Clontech, Palo Alto, Calif.) were cultured in RPMI 1640 medium supplemented with 10% fetal calf serum (FCS), 100 U of penicillin per ml, 100 μg of streptomycin per ml, and 2 mM l-glutamine. U937 and U1 cells overexpressing LIP (U937-LIP and U1-LIP) have been previously described (13, 27, 28) and were cultured under the same conditions. 293T human embryonic kidney cells and 3T3 fibroblasts were grown in an alpha modification of Eagle's medium supplemented with 10% FCS, 100 U of penicillin per ml, 100 μg of streptomycin per ml, and 2 mM l-glutamine.

Macrophage isolation.

Peripheral blood macrophages were isolated from whole blood obtained from healthy HIV-1-seronegative donors. Mononuclear cells were obtained by differential centrifugation using a Ficoll/Hypaque gradient (Sigma, St. Louis, Mo.) as previously described (28). The cells were cultured in RPMI 1640 medium supplemented with 10% FCS, 100 U of penicillin per ml, 100 μg of streptomycin per ml, and 2 mM l-glutamine. Macrophages were separated from lymphocytes by an initial adherence to plastic culture flasks overnight. After removal of nonadherent cells, macrophages were cultured for 5 to 7 days prior to infection.

ICAM-1 and IL-6 antibody blocking.

For blocking of ICAM-1, 1.0 × 10⁵ U1 cells were cocultured with 1.0 × 10⁴ HUVEC in 96-well plates. Mouse immunoglobulin G1 (IgG1) anti-human CD54 monoclonal antibody (MAb) (1 μg/ml; Pharmingen, San Diego, Calif.) was added to the coculture, and viral induction was measured 48 h later by reverse transcriptase (RT) activity. To block IL-6 activity, 1.0 × 10⁵ U1 cells were cultured in a 24-well transwell plate with a 0.4-μm filter (Costar, Corning, N.Y.) with 1.0 × 10⁵ HUVEC. Rat IgG1 anti-human IL-6 MAb (1 μg/ml; Pharmingen) was added to the bottom well containing adherent HUVEC. Viral induction by U1 cells in the top well was measured 48 h later by RT activity. In both cases, mouse IgG1 (1 μg/ml; ICN/Cappel, Aurora, Ohio) was used as an isotype control.

Generation of HIV-1 infectious titers and infections.

Replication competent virus was generated by transfecting 293T cells with 15 μg of M-tropic pHIV-BaL DNA and 3 μg of Rev in a Rous sarcoma virus expression construct (RSV-Rev) DNA by CaPO₄ transfection (40). Wild-type and mutant HXB2 virus were similarly generated with 15 μg of HIV-HXB2 or mC2,C3 HXB2 (27, 28) DNA, and 3 μg of RSV-Rev DNA. Production of viral particles was monitored by RT activity 48 h posttransfection. Infectious virus stocks were added to 1.0 × 10⁶ primary macrophages or U937 monocytes/ml. Equivalent levels of virus were used to infect cells, as determined by RT activity. The infection medium was then removed 24 h postinfection and replaced with fresh RPMI medium supplemented with 10% FCS, 100 U of penicillin per ml, 100 μg of streptomycin per ml, and 2 mM l-glutamine.

Vesicular stomatitis virus glycoprotein (VSV-G)-pseudotyped HIV-1 was also generated by transient transfection using 15 μg of pNL43-Luc(+)Env(−) DNA (12), 3 μg of LVSVG DNA (5), and 3 μg of RSV-Rev DNA. Transfection efficiency was assessed by luciferase activity. For infections with pseudotyped virus, 1 to 2 ml of undiluted viral stocks was added to 1.0 × 10⁶ U937 cells or primary macrophages per ml for 24 to 48 h. Cells were harvested 48 h postinfection and assayed for viral transcription by luciferase assays. Cells (1.0 × 10⁶ to 2.0 × 10⁶) were lysed in 1× reporter lysis buffer (Promega, Madison, Wis.), and supernatants were collected. Cell extract (20 μl) was added to 100 μl of luciferase substrate (Promega), and activity was measured using a luminometer.

Transient transfections of U937 cells.

U937 cells (5.0 × 10⁵ to 6.0 × 10⁵) were transiently transfected with 1.0 μg of LTR-Luc or mC2,3 LTR (26) DNA using Lipofectamine 2000 (Gibco BRL, Rockville, Md.). At 24 h posttransfection, cells were harvested and assayed for luciferase activity.

RT assays.

Virus replication was measured by RT assays at various times postinfection (23). Briefly, 10 μl of supernatant was added to a mixture containing 60 mM Tris, 24 mM dithiothreitol [DTT], 7 mM MgCl₂, 75 mM NaCl, 6 μg of poly(dG)/ml, 12 μg of poly(rC)/ml, 0.06% NP-40, and 10 μCi of [α-³²P]dGTP in a final volume of 50 μl and incubated at 37°C for 1 h. A 10-μl portion of this reaction mixture was then transferred to DEAE paper and washed twice in 2× SSC (0.3 M NaCl, 0.03 M sodium citrate) for 15 min at 25°C. RT activity was quantitated using a phosphorimager. Cells were refed every 3 to 4 days by removing 80% of the spent medium and replacing it with fresh medium.

Immunofluorescent staining.

To detect cell surface adhesion molecules, 1.0 × 10⁶ HUVEC were incubated in staining medium (1% FCS in phosphate-buffered saline [PBS]) with 1 μg of IgG1 isotype fluorescein isothiocyanate (FITC)-conjugated mouse anti-human CD54 MAb (Sigma) or 1 μg of FITC-conjugated mouse IgG2a (Pharmingen) isotype control for 30 min at 4°C. Cells were washed three times in staining medium, and propidium iodide was added at 10 μg/ml. After dead cells had been gated out, fluorescence was measured at 530 nm using a Coulter flow cytometer at the Penn State flow cytometry core facility.

Adhesion assays.

U937 and U937-LIP cells (5.0 × 10⁵) were stained with 1.0 μM 2′,7′-bis-(2-carboxyethyl)-5-(and -6-)carboxyfluorescein-acetoxymethyl ester (BCECF-AM) (Molecular Probes, Eugene, Oreg.) at 37°C for 30 min. Cells were washed twice with PBS and cultured with 5 × 10⁴ HUVEC at 37°C for an additional 30 min in 96-well plates. Cells were washed several times with PBS before fluorescence was measured at 530 nm using a microplate reader.

Nuclear extract preparations and electromobility shift assays (EMSA).

Nuclear extracts from U937 cells were prepared as described previously (49) by lysing 1.0 × 10⁶ to 2.0 × 10⁶ cells with 10% Nonidet P-40 in buffer A (10 mM HEPES [pH 7.9], 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, 0.5 mM phenylmethylsulfonyl fluoride). The extracts were recovered in buffer C (20 mM HEPES [pH 7.9], 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, 1 mM phenylmethylsulfonyl fluoride). Annealed C/EBPβ binding site DNA (50 ng; 5′ GATCGCCTAGCATTTCATCACACGT 3′ and 5′ GATCACGTGTGATGAAATGCTAGGC 3′) or NF-κB binding site DNA (5′ AGCTAAGGGACTTTCCGCTGGGGACTTTCCAGG 3′ and 5′ AGCTCCTGGAAAGTCCCCAGCGGAAAGTCCCTT 3′) were end filled with [α-³²P]dCTP using bacterial Klenow fragment (Promega). The DNA probe was used at a specific activity of 10⁸ to 10⁹ cpm/μg and incubated with 5 μg of nuclear extract samples in a reaction mixture containing 3 μg of poly(dI-dC) (Amersham Pharmacia Biotech, Arlington Heights, Ill.), 0.25 M HEPES (pH 7.5), 0.6 M KCl, 50 mM MgCl₂, 1 mM EDTA, 7.5 mM DTT, and 9% glycerol for 20 min at 25°C. A 50-fold excess of unlabeled C/EBPβ and NF-κB binding site DNA was used as both specific and nonspecific competitors. Anti-C/EBPβ antibody (0.5 μg; Santa Cruz Biotechnology, Santa Cruz, Calif.) was used to supershift complexes. The samples were run on a 6% polyacrylamide gel and visualized by autoradiography.

ELISA.

Enzyme-linked immunosorbent assay (ELISA) plates were coated with 50 μl of either 4-μg/ml mouse anti-human MAb against IL-1β or TNF-α in Ngai's coating buffer (15 mM Na₂CO₃, 34.8 mM NaHCO₃ [pH 9.6]) or 2-μg/ml mouse anti-human MAb against IL-6 in 0.1 M NaHCO₃ (pH 8.2) and incubated overnight at 4°C. After plates had been washed four times in PBS with 0.2% Tween (PBST) for 1 min, plates were blocked with 200 μl of 1% bovine serum albumin–PBS/well and incubated for 2 h at 25°C. Plates were washed again with PBST, and samples were added at 100 μl/well and incubated overnight at 4°C. Plates were washed an additional four times with PBST, and 100 μl of 0.2-μg/ml biotinylated affinity-purified goat anti-human IL-1β, IL-6, and TNF-α (0.1 μg/ml) polyclonal antibodies in 1% bovine serum albumin–PBS were added to each well. Plates were incubated for 2 h at 25°C. After six washes with PBST for 1 min, 100 μl of 1-μg/ml streptavidin peroxidase was added to each well, and samples were incubated for 30 min at 25°C. Plates were washed eight times with PBST for 1 min, and 100 μl of 2,2′-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) (ABTS) substrate solution (Sigma) was added to each well. Samples were incubated 60 to 90 min in the dark at 25°C and read with a microplate reader. All antibodies used for ELISA were purchased from R&D Systems (Minneapolis, Minn.).

RESULTS

Endothelial cells provide a microenvironment that enhances HIV-1 transcription.

Cellular microenvironments created by endothelial and epithelial cells have been demonstrated to influence HIV-1 replication (6, 17, 20, 22, 52). In order to further understand the importance of these cellular interactions, U1 cells were cocultured with HUVEC activated with 10 ng of phorbol myristate acetate (PMA) per ml. U1 cells are promonocytic cells that harbor latent HIV-1 provirus, which can be activated by a variety of factors, including endothelial cell-derived signals (6, 17, 21). The pretreatment of HUVEC leads to the induction of inflammatory cytokines and surface adhesion molecules (58), ensuring optimal interactions between HUVEC and U1 monocytic cells. Consistent with previous observations, when cocultured with stimulated HUVEC, U1 cells produced >25-fold-higher levels of virus than U1 cells cultured alone, as measured by RT activity (Fig. 1A). Furthermore, this ability to enhance virus production in U1 cells was not due to direct effects of PMA treatment and was specific for activated HUVEC, since PMA treatment of empty wells, 293T human embryonic kidney cells, and 3T3 mouse fibroblasts did not result in significant induction of virus replication (data not shown).

FIG. 1.

Endothelial cells enhance HIV-1 replication in monocytes/macrophages through a transcriptional mechanism. (A) U1 cells were cultured alone or with HUVEC pretreated with 10 ng of PMA per ml. Supernatants were assayed for RT activity 48 h after coculture. (B) Primary macrophages (Mφ) were infected with HIV-1 BaL and cultured alone, with HUVEC, or with 3T3 cells. Supernatants were assayed for RT activity 10 days postinfection. (C) U937 cells were infected with VSV-G-pseudotyped HIV-Luc and cultured alone or with HUVEC pretreated with 10 ng of PMA per ml. Cells were lysed 48 h postinfection and assayed for luciferase activity. (D) Primary macrophages were infected with VSV-G-pseudotyped HIV-Luc and cultured either alone or with PMA-stimulated (10 ng/ml) HUVEC. Macrophages were harvested 48 h postinfection and assayed for luciferase activity. These results are from single experiments performed in triplicate and are representative of at least three independent experiments. Error bars show 1 standard deviation.

We also examined the ability of endothelial cells to enhance HIV-1 expression in primary macrophages isolated from peripheral blood. Cells were infected in the absence or presence of endothelial cells with an R5 virus, BaL, and assayed for viral replication by RT activity at various times postinfection. At 10 days postinfection, an eightfold enhancement in viral replication was observed in the primary macrophages cocultured with HUVEC compared to the macrophages cultured alone (Fig. 1B). Enhanced virus expression was detected over the course of 3 weeks postinfection (data not shown). This increase in HIV-1 replication was dependent upon specific interactions between macrophages and endothelial cells, since macrophages cocultured with 3T3 and 293T fibroblast lines did not result in significant changes in virus production (Fig. 1B and data not shown). In addition, HUVEC were not productively infected by HIV-1 (data not shown). These results suggest that signals provided by the endothelial cells enhance HIV-1 replication in primary macrophages.

To address whether endothelial cells enhance HIV-1 expression in macrophages via a transcriptional mechanism, we used VSV-G to pseudotype a replication-incompetent HIV-1 construct in which a luciferase reporter gene (HIV-Luc) was inserted in place of the envelope gene (Env). This VSV-G-pseudotyped virus enters the cell through a receptor-mediated endocytic mechanism, bypassing receptor-mediated fusion and potential signaling events initiated by HIV-1 envelope proteins binding to cellular receptors, thus permitting us to focus specifically on transcriptional effects of cell-cell interactions. U937 monocytic cells infected with HIV-Luc were cocultured with stimulated HUVEC and luciferase activity was measured as an indication of virus transcription. A 14-fold increase in luciferase activity was observed in infected U937 cells cocultured with stimulated HUVEC compared to infected U937 cells cultured alone (Fig. 1C). No significant difference in cell numbers recovered from cultures in the absence and presence of endothelial cells were detected (data not shown). In addition, primary macrophages infected with VSV-G pseudotyped HIV-Luc and cocultured with stimulated HUVEC showed threefold-higher luciferase activity than infected macrophages cultured alone (Fig. 1D), suggesting that endothelial cells mediate the induction of HIV-1 expression in macrophages by a transcriptional mechanism. These results are specific to factors produced by stimulated HUVEC, since pretreatment of 293T or 3T3 cells with PMA did not significantly enhance virus transcription in either U937 cells or primary macrophages (data not shown).

ICAM-1 induces HIV-1 expression in monocytic cells.

Our experiments with primary macrophages demonstrate that direct cell contact between endothelial cells and infected macrophages enhances HIV-1 transcription. Endothelial cells express a number of adhesion molecules that are regulated upon activation, including ICAM-1, which can be strongly induced following PMA treatment (Fig. 2A). Furthermore, previous studies have shown that mature macrophages express both ICAM-1 and its ligand, Mac-1 (46, 57). The importance of ICAM-1 in mediating HIV-1 induction in monocytic cells was tested by blocking ICAM-1 with antibody in U1-endothelial cell cocultures. Addition of anti-ICAM-1 antibody to cocultures of U1 and HUVEC resulted in a 2.4-fold decrease in virus induction compared to cocultures with isotype control (Fig. 2B), suggesting that ICAM-1, in part, provides signals that activate HIV-1 expression in latently infected monocytic cells.

FIG. 2.

ICAM-1 on activated endothelial cells induces HIV-1 expression in monocytic cells. (A) Flow cytometry analysis of HUVEC unstimulated and stimulated with 10 ng of PMA per ml for 24 h. Cells were stained with anti-CD54 (ICAM-1) antibody or mouse IgG isotype matched control. (B) U1 cells were cultured with HUVEC stimulated with 10 ng of PMA per ml. Either anti-ICAM-1 antibody or mouse IgG isotype control (1 μg/ml) was added to the cultures, and RT activity was measured 48 h later. These results are from single experiments performed in triplicate and are representative of four separate experiments. Error bars show 1 standard deviation. ∗, P < 0.05, Student t test.

Endothelial cell-derived IL-6 induces HIV-1 expression in monocytic cells.

Endothelial cells, in addition to being a selective barrier, actively recruit cells and modulate immune responses through the production of soluble factors, including cytokines and chemokines (30, 36, 48, 54). We were interested in identifying cytokines produced by endothelial cells that potentially regulated HIV-1 expression; therefore, supernatants from both unactivated and activated HUVEC were collected and analyzed by ELISA for levels of IL-1β, IL-6, and TNF-α (Table 1). Unstimulated endothelial cells produced low levels of IL-6, but IL-1β and TNF-α were not detected. After stimulation with 10 ng of PMA per ml, IL-6 production was dramatically increased ninefold, whereas IL-1β and TNF-α levels remained low, implying that IL-6 is the primary inflammatory cytokine produced by PMA-stimulated HUVEC.

TABLE 1.

Stimulated HUVEC produce significant amounts of IL-6

Cytokine	Concn (pg/ml) in:		Induction (fold)
	Unstimulated HUVEC	Stimulated HUVECa
IL-1β	0	0	0
IL-6	1,046	9,739	9.3
TNF-α	0	16	16

^aHUVEC were stimulated with 10 ng of PMA per ml for 16 h before supernatants were collected and cytokine levels were measured by ELISA. 

An IL-6-specific antibody was used to block the activity of this cytokine and to determine its role in regulating HIV-1 expression in monocytic cells. U1 cells were cultured in the upper chamber of a transwell apparatus above activated endothelial cells, in the absence or presence of an anti-IL-6 monoclonal antibody or an isotype-matched control. After 48 h, virus production was examined. A 30 to 40% decrease in virus production was consistently observed following treatment with anti-IL-6 antibody compared to the control, suggesting that IL-6 is partly responsible for induction of HIV-1 transcription by endothelial cells (Fig. 3).

FIG. 3.

IL-6 induces HIV-1 expression in monocytic cells. U1 cells were cultured with HUVEC stimulated with 10 ng of PMA per ml in transwells separated with 0.4-μm filters. Anti-IL-6 antibody (1 μg/ml) or mouse IgG isotype control was added to cultures, and RT activity was measured 48 h later. Each data point represents three separate experiments performed in triplicate. Error bars show 1 standard deviation. ∗, P < 0.025, Student t test.

Induction of HIV-1 by endothelial cells requires functional C/EBP factors.

A possible mechanism for the induction of HIV-1 in macrophages by endothelial cell contact is through the activation of transcription factors such as NF-κB and C/EBPβ. Potential target genes for C/EBPβ in monocytic cells include cytokines and adhesion molecules such as ICAM-1 and Mac-1 (2, 3, 27). Therefore, ectopic expression of the C/EBP dominant-negative protein LIP might alter the ability of monocytic cells to bind and adhere to endothelial cells. Monocytic cell lines overexpressing LIP have been previously characterized, and LIP has been demonstrated to specifically inhibit endogenous C/EBP activity in multiple cell types, including monocytic cell lines (26, 27, 28). As shown in Fig. 4, U937 cells were shown to adhere strongly to activated endothelial cells, whereas U937-LIP cells displayed a twofold decrease in their ability to adhere, indicating that endogenous C/EBP factors are required for monocyte-endothelial cell interactions.

FIG. 4.

Functional C/EBP factors required for adherence of monocytes to endothelial cells. U937 and U937 overexpressing LIP cells were stained with 1.0 μM BCECF-AM and cultured alone or with unstimulated or lipopolysaccharide-treated (10 μg/ml; STM) HUVEC for 30 min. Fluorescence was measured at 530 nm to determine adhesion. These results represent single experiments performed in triplicate. Error bars show 1 standard deviation.

Previous studies have indicated that C/EBPβ is strongly upregulated in monocytic cells upon activation and that it is required for HIV-1 transcription and replication in monocytes/macrophages (26, 27, 28). Therefore, we examined whether this transcriptional activator was also required for responses to endothelial cell-derived signals. U937 cells were cocultured with stimulated HUVEC for 24 h, and U937 nuclear extracts were prepared to assay for changes in C/EBP binding activity using EMSA (Fig. 5). After coculture with stimulated endothelial cells, a diffuse complex that specifically bound C/EBP oligonucleotides but not NF-κB sequences was significantly increased in U937 cells (Fig. 5A). Although this complex included multiple C/EBP family members and isoforms, C/EBPβ was the predominant protein in this complex, since antibody specific to C/EBPβ supershifted the induced binding activity (Fig. 5B). In addition to induction of C/EBPβ, we also observed the induction of other factors, including NF-κB (Fig. 5C).

FIG. 5.

Endothelial cells increase binding of C/EBPβ and NF-κB to binding sites from HIV-1 LTR. (A) EMSA was performed with nuclear extracts prepared from U937 cells cultured alone or with HUVEC stimulated with 10 ng of PMA per ml and an oligonucleotide probe spanning the −169 bp C/EBP binding site (27). Extracts were incubated with no competitor (lanes 1 and 2), a 50-fold excess of specific C/EBP binding site competitor (lane 3), and a 50-fold excess of nonspecific NF-κB binding site competitor (lane 4). (B) Extracts were incubated with no antibody (lanes 1 and 2) and 0.5 μg of C/EBPβ antibody (lanes 3 and 4). Complexes did not supershift with 0.5 μg of nonspecific control, NF-κB, antibody (data not shown). (C) Extracts were probed with an NF-κB binding site oligonucleotide and incubated with no competitor (lanes 1 and 2), a 50-fold excess of specific NF-κB binding site competitor (lane 3), and a 50-fold excess of nonspecific C/EBP binding site competitor (lane 4).

To specifically investigate the requirement for C/EBP factors in endothelial cell-induced HIV-1 transcription in monocytic cells, U1 cells and U1 cells overexpressing LIP (U1-LIP) were cocultured with HUVEC. Endothelial cells activated with PMA strongly induced HIV-1 replication, producing approximately 100-fold more virus than U1 cells cultured alone. However, this induction was decreased by approximately fourfold in U1-LIP cells (Fig. 6A), indicating that endogenous C/EBP factors are necessary for efficient activation of HIV-1 transcription by endothelial cells. The observed induction of viral replication is specific to endothelial cells since U1 coculture with 293T cells resulted in only modest virus production (data not shown).

FIG. 6.

Functional C/EBP factors required for endothelial cell-mediated induction of HIV-1 in monocytic cells. (A) U1 and U1-LIP cells were cultured alone or with HUVEC stimulated with 10 ng of PMA per ml. RT activity was measured after 48 h. These results are representative of three independent experiments. (B) U937 and U937-LIP cells were infected with VSV-G pseudotyped HIV-Luc and cultured alone or with HUVEC stimulated with 10 ng of PMA per ml. U937 and U937-LIP cells were harvested 48 h postinfection and assayed for luciferase activity. These results are from single experiments performed in triplicate and are representative of three separate experiments. (C) U1 and U1-LIP cells were cultured alone or with PMA-stimulated (10 ng/ml) HUVEC in transwells separated by 0.4-μm filters. RT activity was measured 48 h later. These results are representative of three independent experiments. Error bars show 1 standard deviation.

To determine if the inhibition of HIV-1 expression observed in Fig. 6A was through a transcriptional mechanism, U937 and U937-LIP cells were infected with VSV-G-pseudotyped HIV-Luc for 24 h before coculturing with stimulated HUVEC. In cocultures with stimulated endothelial cells, a sixfold increase in luciferase activity was detected in U937 cells compared with U937 cells cultured alone (Fig. 6B), whereas HIV-1 transcription was reduced by approximately 75% in U937-LIP cells cultured with stimulated HUVEC. Taken together, these data indicate that C/EBPβ activity is induced in monocytes/macrophages upon interacting with stimulated endothelial cells and that endogenous C/EBP factors are necessary for efficient induction of HIV-1 in response to these interactions. Furthermore, the decrease in virus transcription upon overexpression of LIP is not due to indirect effects of LIP on other transcription factors, since U937-LIP cells induced NF-κB binding activity following coculture with PMA-treated endothelial cells as well as control U937 cells cocultured with activated endothelial cells (data not shown).

The potential role of C/EBPβ in the induction of HIV-1 by endothelial cell-derived cytokines was examined by separating monocytic cells from HUVEC using a transwell system and measuring virus production. A 10-fold induction in virus replication was observed in U1 cells 48 h after culture with stimulated endothelial cells in the lower chamber (Fig. 6C). The lack of significant virus production when U1 cells were cultured with stimulated 293T cells suggests that the induction of viral replication was due to specific soluble factors produced by HUVEC and not residual PMA from pretreatment of the cells (data not shown). This induction of HIV-1 expression requires functional C/EBP factors, since U1-LIP cells did not produce significant levels of virus when cultured with stimulated endothelial cells in the transwell (Fig. 6C), indicating that these factors are necessary for monocytic cells to fully respond to endothelial cell-produced cytokines.

Endothelial cell-induced HIV-1 transcription is dependent upon C/EBP binding sites in the HIV-1 LTR.

The experiments described above indicate that C/EBP proteins are required for endothelial cell-mediated induction of HIV-1 transcription. Previous studies have demonstrated that there are two critical C/EBP sites within the HIV-1 LTR, C2, located at −178 to −159, and C3, positioned at −120 to −109 (26, 27, 59). To directly assess the importance of these C/EBP binding sites, HIV-1 LTR-Luc reporter constructs with either wild-type high-affinity C/EBP sites or mutated C/EBP sites (mC2,3 LTR) (26) were transiently transfected into U937 cells and cultured either alone or with PMA-stimulated endothelial cells for 24 h before luciferase activity was examined. As shown in Fig. 7A, HIV-1 LTR activity was induced 2.5- to 3-fold in cocultures with endothelial cells. However, no significant induction of LTR activity by endothelial cells was observed with the mC2,3 LTR construct, demonstrating the necessity for C/EBP sites in this response.

FIG. 7.

Induction of HIV-1 transcription and replication by endothelial cells requires C/EBP binding sites in HIV-1 LTR. (A) U937 cells were transiently transfected with 1.0 μg of LTR-Luc or mC2,3 LTR DNA and cultured alone or with HUVEC pretreated with 10 ng of PMA per ml. Luciferase activity was measured 24 h posttransfection. These data are from two experiments performed in triplicate. Error bars show 1 standard deviation. (B) U937 cells were infected with equivalent levels of wild-type HIV-1 HXB2 or mutant mC2,C3 HXB2 virus and cultured with PMA-treated (10 ng/ml; STM) HUVEC. Supernatants were assayed for RT activity 4 days postinfection. These results are from a single experiment performed in triplicate and are representative of four separate experiments. Error bars show 1 standard deviation. ∗, P < 0.03, Student t test.

HIV-1 clones harboring mutations in the high-affinity C/EBP sites were used to confirm the importance of these sequences in regulating HIV-1 expression in response to endothelial cell signals. Previous studies have shown a requirement for C/EBP sites during HIV-1 replication in monocytes/macrophages (27, 28) (data not shown). To specifically address the role of C/EBP sites in HIV-1 replication in the presence of endothelial cells, U937 cells were infected with wild-type HXB2 virus or virus with mutated C/EBP sites (mC2,C3 HXB2) (27, 28) and cultured with PMA-treated endothelial cells. As shown in Fig. 7B, the ability of mC2,C3 HXB2 to replicate in response to endothelial cell signals was reduced by 60% compared to that of wild-type HXB2. These data indicate that C/EBP sites within the HIV-1 LTR are necessary for efficient virus replication in monocytes/macrophages in response to microenvironmental signals.

DISCUSSION

Previous studies have indicated that tissue microenvironments influence HIV-1 replication (6, 17, 20, 22, 52). Although it has been suggested that endothelial cells might influence HIV-1 replication by inducing factors such as NF-κB, the role of other transcriptional activators has not been fully investigated. Our results suggest that, through the production of soluble factors and direct cell contact, endothelial cells provide a favorable environment for HIV-1 transcription in macrophages and extend previous studies by demonstrating that this induction of virus transcription requires functional C/EBPβ.

Upon coculture with endothelial cells, an enhancement of HIV-1 replication and transcription in monocytic cell lines and primary macrophages was observed. Although we show that the induction of HIV-1 in monocytic cells depends on specific cellular interactions with endothelial cells, the ability to transcriptionally activate HIV-1 is not unique to endothelium. Lung epithelial cells have also been shown to enhance virus expression through similar mechanisms, including the activation of transcription factors such as C/EBPβ and NF-κB (data not shown) (10). Taken together, these results support models proposing that specific tissue microenvironments can alter the course of infection at different sites, including the brain, lymph nodes, and lungs, by promoting HIV-1 transcription.

Although other reports have observed effects in cocultures between monocytic cells and unstimulated endothelial cells (6, 17), we detected moderate induction of virus expression with unstimulated endothelial cells and needed to activate endothelial cells to consistently induce robust virus expression. Stimulating endothelial cells with PMA, lipopolysaccharide, or inflammatory cytokines dramatically upregulates expression of ICAM-1 and inflammatory molecules such as IL-6, which we and others have shown are critical for mediating macrophage-endothelial cell interactions and transcriptional activation of HIV-1 expression in monocytic cells (6, 17, 30, 52, 54, 58). Our studies further corroborate these findings by indicating an increased level of ICAM-1 on stimulated endothelial cell surfaces and demonstrating that antibody to ICAM-1 decreases endothelial cell-induced HIV-1 replication in U1 cells.

Inflammatory cytokines from endothelial cells also influence the expression of HIV-1 in macrophages (15, 21, 33, 39, 41, 42). Previous studies have shown that IL-1β, IL-6, and TNF-α levels are elevated in the sera of AIDS patients (7, 19, 34). Furthermore, IL-6 has been shown to increase virus expression and have a role in promoting AIDS-associated diseases, including Kaposi's sarcoma and non-Hodgkin's lymphoma (18, 19, 36, 41). Our results, which demonstrate that endothelial cell-produced IL-6 increases HIV-1 transcription, are consistent with IL-6 having an important role in regulating HIV-1. In addition, other endothelial factors are clearly involved in regulating HIV-1 replication in monocytic cells, since blocking IL-6 or ICAM-1 does not completely inhibit HIV-1 expression.

More importantly, our data demonstrate a role for C/EBPβ in endothelial cell-induced HIV-1 expression. When C/EBP factors are functionally inhibited by a dominant-negative regulator (LIP), endothelial cell-mediated induction of HIV-1 proviral transcription is decreased. Furthermore, full induction of LTR activity or virus replication requires C/EBP binding sites. This enhancement of virus expression was inhibited by LIP overexpression, indicating that C/EBP factors regulate HIV-1 expression through a transcriptional mechanism. The ability of LIP to inhibit virus expression was not due to a general block in transcription factor activity, since NF-κB binding activity was similar in both U937 and U937-LIP cells following coculture with endothelial cells. C/EBPβ most likely regulates HIV-1 transcription, since this activator was induced in monocytic cells upon coculture with stimulated endothelial cells. These results showing that functional C/EBPβ is required for HIV-1 transcription in monocytes/macrophages are consistent with previous studies demonstrating the importance of C/EBPβ for HIV-1 transcription and replication (26, 27, 28, 47, 59).

C/EBPβ has been shown to be necessary but not sufficient for transcriptional activation of the HIV-1 LTR. Overexpression of C/EBPβ alone does not induce latent proviral expression, indicating that posttranslational events or protein-protein interactions are required for appropriate HIV-1 induction (27). Recent studies demonstrate that C/EBPβ can interact both physically and functionally with the basal transcriptional machinery as well as multiple transcription factor families (1, 51, 61) and chromatin remodeling complexes (16, 37, 53, 56, 60). In particular, C/EBPβ can interact with NF-κB to synergistically activate the HIV-1 LTR (47). The potential interaction between C/EBPβ and NF-κB, which are both induced by endothelial cells, could compensate for the mutated C/EBP binding sites and contribute to the modest reduction in HIV-1 transcription and replication. Combinatorial mechanisms between various transcription factors and coactivators are clearly required for appropriate induction and expression of the HIV-1 LTR; however, our data are consistent with the conclusion that C/EBPβ is a critical component of the transcriptional activation complex found in macrophages.

Although C/EBPβ is required for HIV-1 transcription in macrophages, it may also indirectly influence HIV-1 replication by altering the interactions between macrophages and endothelial cells. One functional consequence of inhibiting endogenous C/EBP by ectopic LIP expression in monocytic cells is a decrease in the ability of cells to bind to endothelial cells. These results indicate that C/EBPβ is required for expression of genes that mediate monocyte-endothelial cell interactions. C/EBPβ expression is induced upon macrophage activation and differentiation (38) and has been shown to coordinate the expression of many genes involved in macrophage activation, such as ICAM-1 and Mac-1 (27) and inflammatory cytokines (2; reviewed in references 45 and 61), all of which have functionally important NF-κB and C/EBPβ sites within their promoters (11, 35, 58). Furthermore, we have preliminary data suggesting that IL-6 gene expression in macrophages requires functional C/EBPβ. In this study, we demonstrate the importance of microenvironments in regulating HIV-1 transcription in macrophages, particularly through endothelial cell-mediated signals. One such signal, IL-6, can further increase C/EBPβ expression (1), exemplifying the autoregulatory roles of IL-6 and C/EBPβ. Together, these observations underscore the importance of C/EBPβ in regulating macrophage function and coordinating gene expression during the inflammatory response initiated by HIV-1 infection.