INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

THE MECHANISM OF SPREAD of human immunodeficiency virus (HIV) infection to adjacent cells or tissues is not completely understood. It has been concluded that soluble proteins likely contribute to the early stages of infection. One candidate molecule is the HIV-1 viral protein known as the transactivator of transcription (HIV-1 Tat). This protein is believed to play a critical role in the progression of HIV disease (11).

The HIV-1 Tat protein released by HIV-infected monocytes is believed to be biologically active. Tat has been identified in brain and other tissue of HIV- infected patients (37, 39). The mechanism by which Tat protein is expressed in nonmonocytic cells during infection is not understood. Cellular uptake of Tat protein, a potentially important part in this process, has been documented in neuronal cells and astrocytes (15, 22). The cellular uptake of Tat protein during HIV infection is believed to modify gene expression within adjacent cells. The alteration of gene expression in parenchymal cells has become an area of heightened interest to delineate a possible mechanism to explain the progression of acquired immunodeficiency syndrome (AIDS) infection and its complications and even primary vascular disease, atherosclerosis, and hypertension associated with AIDS infection (34).

It is likely (but not fully tested) that activation of endothelial cells is an early event leading to monocyte adhesion, transendothelial migration of HIV-1-infected monocytes into adjacent parenchyma, and subsequent infection of tissue. We believe that HIV-1 Tat protein is a key player in this initial stage. Indeed, exogenous HIV-1 Tat protein has been shown to induce expression of vascular cell adhesion molecule-1 and intercellular adhesion molecule-1 in human astrocytes (40). Vascular cell adhesion molecule-1 and intercellular adhesion molecule-1 are also upregulated in transgenic mice expressing the envelope protein HIV-1 gp120 (35).

Previous studies indicated that exogenous HIV-1 Tat protein increases expression of E-selectin and enhances production of interleukin-6 and interleukin-8 in cultured human endothelial cells (12-14). These actions of exogenous Tat protein might lead to activation of endothelial cells and increased endothelial cell permeability and might even facilitate the development of Kaposi's sarcoma, an aggressive vascular tumor frequently found in AIDS-infected patients (1). Indeed, HIV-1 tat gene expression in transgenic mice is known to induce endothelial cell proliferation and tumors resembling Kaposi's sarcoma (7). Transgene expression of HIV Tat is also known to induce cardiomyopathy (26), suggesting that cardiovascular complications may involve more than just vascular complications. Adhesion molecule gene expression and enhanced monocyte adhesion may lead to accelerated atherosclerosis. Indeed, there is emerging evidence that HIV-1 infection is a risk factor for vasculopathy and atherosclerotic disease (4, 20). It is not understood whether this is due to drug treatment or an underlying complication of infection. The strong association of monocyte adhesion with atherosclerosis and the stimulation of cell adhesion molecule gene expression by exogenous Tat protein suggest that viral proteins may play a significant role in the development of vascular disease.

Despite these findings, no information is available regarding the proximal signaling events in the regulation of cell adhesion to endothelial cells activated by HIV-1 Tat. There is ample evidence that activation of the transcription factor nuclear factor B (NF-B) plays a significant role in cytokine-stimulated cell adhesion molecule expression in endothelial cells owing to known binding sites for NF-B in the promoter regions for genes encoding cell adhesion molecules (27). Although exogenous Tat protein is known to activate NF-B in immune cells, including monocytes and T lymphocytes (8, 17, 24), it is not known whether exogenous Tat or transfection to express endogenous tat is able to activate this pathway in endothelial cells.

To our knowledge, this is the first report documenting the role of NF-B activation in initiation of monocyte adhesion to human endothelial cells transfected with pcTat.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Endothelial cells and culture media. Human umbilical vein endothelial cells were obtained from Clonetics (Walkersville, MD). Cells were initially cultured in 75% endothelial basal medium (EBM, Clonetics) supplemented with EGM-2 growth components + 25% RPMI 1640 (per company specifications). Confluent cells were passaged in 50% EBM + 50% RPMI 1640, converted to 25% EBM + 75% RPMI, and finally converted to 100% RPMI 1640 containing L-glutamine, fibroblast growth factor, and 5% fetal bovine serum. Media replacement was designed to minimize potential complicating components in EBM and to achieve a physiological glucose concentration of 5.5 mM to prevent activation of NF-B due to elevated glucose concentrations, which we have shown can independently activate NF-B (25). Cells were cultured up to passages 5-7.

Transfections and reporter assays. Optimal transfection conditions for 24 or 48 h were determined using various concentrations of Lipofectin and green fluorescence protein plasmid (Promega, Middleton, WI). Fluorescence was viewed on a fluorescence microscope (Olympus, Lake Success, NY). A plasmid construct, pcTat, was used to transfect endothelial cells with HIV-1 tat gene. For control purposes, we used p12S,FS empty vector. For the reporter assays, endothelial cells were cotransfected with 0.3-1.0 µg of the NF-B-dependent luciferase vector pNF-B-Luc (Clontech Labs, Palo Alto, CA) and compared with the NF-B null vector pTA-Luc (Clontech Labs). Cells were transfected for 48 h with RPMI 1640 as described above. Normalized luciferase activity was determined using an E4030 luciferase assay system and reporter lysis buffer (Promega, Madison, WI).

Electrophoretic mobility shift assay. Nuclear extracts were prepared, and NF-B binding activity was analyzed by electrophoretic mobility shift assay after the extracts were run on 4% gels using a consensus NF-B oligonucleotide (Promega, Madison, WI), as previously described in detail (6, 28).

Prevention of NF-B activation. Cells were incubated with SN-50 peptide (30 µg/ml; Biomol, Plymouth Meeting, PA) during transfection with pcTat. This peptide contains the nuclear localization sequence for NF-B dimers linked to a cell-permeable motif of Kaposi's growth factor and functions to block nuclear translocation of NF-B dimers. In other studies, cells were transfected with pCMV4, IB, S32/36A, a dominant-negative mutant IB plasmid recently denoted the superrepressor IB (provided by Dr. Warner C. Greene, Gladstone Institute of Virology and Immunology, University of San Francisco, San Francisco, CA). This plasmid, in which alanine is substituted at serine residues 32 and 36, does not allow phosphorylation of IB and subsequent release and translocation of NF-B dimer subunits (32). To assess a role of reactive oxygen, some cells were incubated with 10 mM aspirin. This high concentration of aspirin possesses antioxidant efficacy and has been shown previously to block NF-B activation and monocyte adhesion to human umbilical vein endothelial cells stimulated with tumor necrosis factor- (36).

Monocyte adhesion. U-937 monocytic cells were cultured in RPMI 1640 containing 10 mM HEPES, 1 mM sodium pyruvate, 2 mM L-glutamine, 4.5 g/l glucose, and 10% fetal bovine serum. Cell adhesion was analyzed by fluorescence labeled with 50 µg of 2',7'-bis(2-carboxyethyl)-5(6)-carboxyfluorescein acetoxymethyl ester (Molecular Probes, Eugene, OR). Washed-labeled U-937 cells were counted (Coulter, Opa Locka, FL) and resuspended in medium at 5 × 10⁵ cells/ml. A 1.0-ml suspension of U-937 cells was added to transfected endothelial cells and incubated for 37°C for 30 min. Plates were rinsed three times, and 250 µl of lysis buffer (i.e., 1 mM Tris buffer, pH 8.0, and 1% Triton X-100) were added to each well. Samples were transferred to a black DynaTech 96-well plate and read on a CytoFluor II fluorescent plate reader (PerSeptive Biosystems, Foster City, CA) at 485 nm (excitation) and 530 nm (emission). Fluorescence was calibrated against 2',7'-bis(2-carboxyethyl)-5(6)-carboxyfluorescein acetoxymethyl ester-labeled U-937-derived standards.

Statistical analysis. Values are means ± SE. ANOVA or Student's t-test, as appropriate, was used for statistical analysis. P < 0.05 was selected to denote statistical significance.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Electrophoretic mobility shift assay indicated increased NF-B binding activity in nuclear extracts of endothelial cells transfected with pcTat (Fig. 1, lanes 5 and 6) compared with control DNA (lane 7). The increase in NF-B binding activity was greater than that achieved by transfection with the human leukemia virus Tax protein (lanes 3 and 4) or by acute stimulation for 1 h with 100 ng/ml phorbol 12-myristate 13-acetate (PMA; lane 8). The specificity for NF-B was verified using 100-fold excess cold wild-type or mutant oligonucleotide in the assay reaction mixture (lanes 1 and 2).

View larger version (38K): [in this window] [in a new window]   Fig. 1.   Electrophoretic mobility shift assay showing increased nuclear factor-B (NF-B) binding activity in nuclear extracts of human umbilical vein endothelial cells (HUVECs) transfected for 48 h with pcTat (lanes 5 and 6) compared with control cDNA (lane 7), transfection with pcTax (lanes 3 and 4), or acute stimulation for 1 h with 100 ng/ml phorbol 12-myristate 13-acetate (PMA) without pcTat (lane 8). Specificity for NF-B is confirmed using 100-fold excess cold wild-type (WT) or mutant (Mut) oligonucleotide (lanes 1 and 2, respectively).

These studies were complemented by functional NF-B activation analysis using NF-B-linked luciferase reporter assays. pcTat was shown to produce a pronounced increase in luciferase activity (Fig. 2). The magnitude of this increase was similar to that achieved by an acute stimulation with a potent concentration of 100 ng/ml PMA in the absence of pcTat. As a control, pcTat produced no activation in cells transfected with a vector lacking NF-B elements, pTA-Luc.

View larger version (28K): [in this window] [in a new window]   Fig. 2.   Stimulation of NF-B-dependent transcriptional activity in endothelial cells cotransfected with 1.0 µg of pNF-B-Luc and 0.3 µg of pcTat. Magnitude of activation is compared with acute (1 h) stimulation with 100 ng/ml PMA without pcTat. Results in relative fluorescence (light) units (RLU) are normalized to protein content and compared with cells contransfected with pTA-Luc, a luciferase reporter plasmid lacking NF-B sequences. Values are means ± SE of 4 different determinations each. P < 0.01 vs. respective pTA-Luc only; ()P < 0.01 vs. respective pNF-B-Luc only.

Functional NF-B activation was also determined in experiments using the HIV-LTR-CAT reporter assay system. Transfection with pcTat caused a potent activation of NF-B-dependent CAT activity compared with control cDNA (Fig. 3). pcTat produced a calculated 28-fold increase in CAT activity compared with only a 3-fold increase in CAT activity produced by pcTax (Fig. 3). CAT activity was markedly lower (i.e., only a 13-fold increase) in pcTat-stimulated cells cotranfected with mutant pB-HIV-CAT. In contrast to pcTat, pcTax did not stimulate CAT activity in cells transfected with pB-HIV-CAT.

View larger version (97K): [in this window] [in a new window]   Fig. 3.   Autoradiogram showing chloramphenicol acetyltransferase (CAT) activity in HUVECs cotransfected with HIV-LTR-CAT (containing normal NF-B sequences) or mutant (Mut) B HIV-LTR-CAT (lacking NF-B sequences) and stimulated with pcTat or with pcTax for comparison. Identity of spots is as follows: nonacetylated (bottom row), monoacetylated (middle row), and diacetylated (top row) species.

Finally, luciferase reporter assays were used in studies with cotransfection with mutant IB S32/36A. Transfection with mutant IB completely blocked pcTat-induced NF-B luciferase activity (Fig. 4). As a positive control, similar findings were obtained in PMA-stimulated cells. As an additional control performed originally in PMA-stimulated cells, using a control empty vector, we determined that dominant-negative mutant IB blocked NF-B luciferase activity by 94.6 ± 0.8% compared with only marginal changes (12.2 ± 2.2% inhibition, n = 2).

View larger version (21K): [in this window] [in a new window]   Fig. 4.   pcTat-induced NF-B transcriptional activity in endothelial cells is abrogated by cotransfection with 1.0 µg of dominant-negative mutant IB S32/36A (superrepressor). Results are compared with PMA-stimulated activity as a positive control. P < 0.05 vs. pcTat or PMA alone.

Endothelial cells that were transfected with pcTat showed increased monocyte adhesion compared with Lipofectin alone (Fig. 5) or compared with a control DNA vector lacking NF-B activity (not shown). The increase in monocyte adhesion stimulated by endogenous tat gene was inhibited by 50% when cells were coincubated with SN-50 peptide (Fig. 6) and completely blocked by cotransfection with the dominant-negative mutant or IB superrepressor (Fig. 5).

View larger version (12K): [in this window] [in a new window]   Fig. 5.   Transfection of endothelial cells with pcTat stimulates monocyte adhesion compared with Lipofectin control. Cotransfection with dominant-negative mutant IB S32/36A (superrepressor) blocks pcTat-induced monocyte adhesion. P < 0.01 vs. Lipofectin and vs. pcTat + mutant IB (superrepressor).

View larger version (9K): [in this window] [in a new window]   Fig. 6.   Monocyte adhesion to pcTat-stimulated endothelial cells (n = 3 determinations each) is attenuated by incubation with 30 µg/ml SN-50 peptide (n = 3 each). *P < 0.05 vs. pcTat.

Incubation with 10 mM aspirin blocked the increase in NF-B luciferase activity by 90% in endothelial cells stimulated with pcTat (Fig. 7). In addition, aspirin normalized the increase in monocyte adhesion in pcTat-stimulated endothelial cells (Fig. 8).

View larger version (14K): [in this window] [in a new window]   Fig. 7.   NF-B luciferase activity in pcTat-stimulated endothelial cells is blocked in presence of 10 mM aspirin. *P < 0.01 vs. pcTat without aspirin (n = 4 determinations each).

View larger version (13K): [in this window] [in a new window]   Fig. 8.   Monocyte adhesion to pcTat-stimulated endothelial cells is normalized to control DNA by 10 mM aspirin. *P < 0.01 vs. control DNA transfection (n  3 determinations each).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The new findings in this study are that endogenous expression of tat gene stimulated monocyte adhesion to human endothelial cells and that this process occurred via a redox-sensitive, NF-B-dependent pathway. Recently, one laboratory has performed peripheral measurements that indicate significantly increased levels of plasma soluble cell adhesion molecules in HIV-positive patients (5, 31). In addition, increased cell adhesion to endothelium has been reported in HIV-1-infected patients. These observations indirectly suggest activation of endothelial cells in vivo but do not prove a causal link, the signaling events involved in endothelial cell activation, or the specific involvement of HIV-1-derived soluble factors, including viral proteins.

A clue to the potential direct actions of HIV-1 viral proteins is provided by studies that show increased interleukin-6 production and increased expression of E-selectin, a key cell adhesion molecule, as a result of incubation of human endothelial cells with exogenous Tat protein (12-14). The molecular events involved in signaling of gene expression by HIV-1 Tat within endothelial cells are unknown. Activation of NF-B is a possible pathway that has not been examined in detail. We hypothesized that this pathway was important, because the genes for cell adhesion molecules contain an NF-B binding site in the promoter region for these genes. A role for NF-B activation in HIV-1 is indirectly supported by a report of increased immunoreactive NF-B in brains of children with HIV-1 encephalitis (10). Also, exogenous HIV-1 Tat has been shown to cause activation of NF-B in Jurkat T cells (2, 29, 38) and in macrophages (3). Despite these findings, the potential involvement of HIV-1 Tat-induced activation of NF-B specifically within endothelial cells per se and the association with increased monocyte adhesion were not previously known. Thus the present study provides novel information in this area.

To understand the process of Tat-induced activation of endothelial cells more fully, we elected to examine the signaling events due to endogenous tat gene expression to eliminate the confounding actions of exogenous Tat protein interaction with plasma membrane receptors that might also induce NF-B activation. To our knowledge, our findings are the very first report of human endothelial cell activation of NF-B and subsequent monocyte adhesion by endogenous expression of tat.

We have documented NF-B activation by several criteria. First, gel shift assays confirmed increased NF-B binding activity after transfection with pcTat. Studies conducted with control cDNA confirm that this increase is due to pcTat and not simply to vector transfection in general. Second, NF-B activation was also confirmed in protocols designed to assess functional NF-B activity using NF-B-dependent luciferase reporter assays. In this instance, pcTat produced a pronounced increase in luciferase activity that was equipotent to PMA, a well-known stimulant of NF-B activation, within a variety of endothelial cells. Third, luciferase activity was not enhanced in pcTat-transfected endothelial cells that were cotransfected instead with the pTA-Luc reporter that lacks B elements. These findings indicate that NF-B activation by pcTat cannot be accounted for by the transfection process alone. Fourth, the pcTat-induced increase in NF-B luciferase activity was completely blocked by cotransfection to express the dominant-negative mutant IB or superrepressor. This mutant contains point mutations for critical serine residues that prevent phosphorylation of IB, which in turn would prevent dissociation of the active NF-B dimer subunits and subsequent nuclear translocation and DNA binding. These findings imply that IB phosphorylation is required for pcTat-induced activation of endothelial cell NF-B. These latter findings represent yet another independent confirmation that pcTat stimulates functional NF-B-dependent transcriptional activity.

Finally, we performed additional studies of functional NF-B activation using the pHIV-1-LTR-CAT reporter, which is used to assess potential for viral replication. HIV-1 Tat is believed to cause transactivation of viral gene expression via binding of Tat protein to stem-loop structures denoted as transactivation-responsive (TAR) elements (16, 23). Our studies indicate marked increases in CAT activity by pcTat but not by a vector control DNA. Unlike endogenous expression of HTLV-1-Tax protein, endogenous tat gene was a more potent stimulant of HIV-1-LTR-CAT reporter activity. We observed that a significant portion of this activity was NF-B dependent, because pcTat produced diminished activity via transfection with a mutant plasmid construct that lacked B sequences.

Interestingly, a lesser but significant amount of CAT activity remained after stimulation with pB-HIV-CAT, a mutated B reporter of pcTat. This finding suggests that endogenous tat may also cause transactivation via an NF-B-independent, TAR-dependent transactivation. In other cell lines, it is known that Tat can regulate gene expression via TAR-dependent and TAR-independent pathways (33). A conclusion from these studies is that a significant portion of viral replication of HIV-1 within endothelial cells, as denoted in the HIV-LTR-CAT vs. pB-HIV-CAT assays, may not require TAR interaction. Accordingly, our studies are new, in that they describe NF-B-independent and NF-B-dependent components of HIV-1 transactivation within human vascular endothelial cells.

Another new finding of our study is that transfection of normal human endothelial cells with pcTat results in activation of endothelial cell NF-B and enhanced monocyte adhesion. Previously, it was shown that HIV infection of monocytes stimulated adhesion to human endothelial cells (9). Additional studies revealed that incubation with extracellular HIV-1 Tat protein is capable of activating purified monocytes and, in turn, stimulating monocyte adhesion to unstimulated endothelium (19). Our new findings using the pcTat vector indicate that HIV-1 Tat protein may also directly activate endothelial cells to facilitate monocyte adhesion via a mechanism involving activation of endothelial cell NF-B. That HIV-1 Tat-activated monocyte adhesion occurs via an NF-B-dependent mechanism was verified in our studies showing that cotransfection with dominant-negative mutant IB under conditions that block NF-B-linked reporter activity also blocks monocyte adhesion to pcTat-stimulated endothelial cells.

Additional implications and potential limitations for protein transduction. Because of the known efficient uptake of Tat protein by various cell types, there has been a heightened focus on the potential use of fusion proteins to Tat protein for protein transduction (18, 30). Because of potential limitations of adenoviral transfection, development of protein transduction technology is a novel pharmacological alternative for the delivery of proteins, compounds, and DNA to cells. The present study using transfection with pcTat suggests caution in the use of this potentially exciting pharmacological approach. Accordingly, future studies are warranted to more completely understand the actions and impact of truncated and full-length intracellular HIV-1 Tat protein on endothelial cell activation, gene expression, and function.