Effects of HIV-1 gp120 and TAT-derived microvesicles on endothelial cell function

Abstract

The aims of this study were twofold. The first was to determine if human immunodeficiency virus (HIV)-1 glycoprotein (gp) 120 and transactivator of transcription (Tat) stimulate the release of endothelial microvesicles (EMVs). The second was to determine whether viral protein-induced EMVs are deleterious to endothelial cell function (inducing endothelial cell inflammation, oxidative stress, senescence and increasing apoptotic susceptibility). Human aortic endothelial cells (HAECs) were treated with recombinant HIV-1 proteins Bal gp120 (R5), Lav gp120 (X4), or Tat. EMVs released in response to each viral protein were isolated and quantified. Fresh HAECs were treated with EMVs generated under control conditions and from each of the viral protein conditions for 24 h. EMV release was higher (P < 0.05) in HAECs treated with R5 (141 ± 21 MV/µl)_, X4 (132 ± 20 MV/µl), and Tat (130 ± 20 MV/µl) compared with control (61 ± 13 MV/µl). Viral protein EMVs induced significantly higher endothelial cell release of proinflammatory cytokines and expression of cell adhesion molecules than control. Reactive oxygen species production was more pronounced (P < 0.05) in the R5-, X4- and Tat-EMV-treated cells. In addition, viral protein-stimulated EMVs significantly augmented endothelial cell senescence and apoptotic susceptibility. Concomitant with these functional changes, viral protein-stimulated EMVs disrupted cell expression of micro-RNAs 34a, 126, 146a, 181b, 221, and miR-Let-7a (P < 0.05). These results demonstrate that HIV-1 gp120 and Tat stimulate microvesicle release from endothelial cells, and these microvesicles confer pathological effects on endothelial cells by inducing inflammation, oxidative stress, and senescence as well as enhancing susceptibility to apoptosis. Viral protein-generated EMVs may contribute to the increased risk of vascular disease in patients with HIV-1.

NEW & NOTEWORTHY Human immunodeficiency virus (HIV)-1-related proteins glycoprotein (gp) 120 and transactivator of transcription (Tat)-mediated endothelial damage and dysfunction are poorly understood. Endothelial microvesicles (EMVs) serve as indicators and potent mediators of endothelial dysfunction. In the present study we determined if HIV-1 R5- and X4-tropic gp120 and Tat stimulate EMV release in vitro and if viral protein-induced EMVs are deleterious to endothelial cell function. gp120 and Tat induced a marked increase in EMV release. Viral protein-induced EMVs significantly increased endothelial cell inflammation, oxidative stress, senescence, and apoptotic susceptibility in vitro. gp120- and Tat-derived EMVs promote a proinflammatory, pro-oxidative, prosenescent, and proapoptotic endothelial phenotype and may contribute to the endothelial damage and dysfunction associated with gp120 and Tat.

INTRODUCTION

Human immunodeficiency virus (HIV)-1 infection is associated with an increased risk and prevalence of atherosclerotic cardiovascular disease (ASCVD) (25). The pathogenesis of HIV-1-associated atherosclerosis is complex (17, 33). Traditional risk factors and exposure to antiretroviral therapy do not fully account for the elevated risk and increased incidence and severity of atherosclerotic vascular disease associated with HIV-1 (25). Endothelial cell activation, inflammation, dysfunction, and death are primary initiating factors in the etiology of atherosclerosis (29, 35) and are associated with HIV-1 infection (17, 33). Importantly, however, the direct and indirect atherogenic effects of HIV-1 on endothelial cell biology are diverse and not completely understood (17, 33). HIV-1-related proteins, such as envelope glycoprotein (gp)120 and transactivator of transcription (Tat), are thought to play a role. Released by HIV-1-infected monocytes, these proteins have been shown to disrupt endothelial cell function and jeopardize survival (18, 33, 42). For example, gp120 is cytotoxic, activating pro-oxidative and proapoptotic pathways (18, 40), and Tat has been shown to induce endothelial inflammation, monocyte adhesion, and apoptosis (23). Viral protein-induced microvesicle release from endothelial cells may represent an additional mediator of endothelial dysfunction with HIV-1.

It is now recognized that a primary indicator of endothelial cell activation, injury, and death is the vesiculation and release of endothelial microvesicles (11). Endothelial cell-derived microvesicles (EMVs) are small (0.1–1 µM in diameter) extracellular anucleod vesicles that, when released, can serve as both a consequence and cause of endothelial dysfunction and as a biomarker and predictor of vascular health and disease (5, 11, 45). Clinically, elevated circulating levels of EMVs have been reported in several cardio/cerebrovascular-related diseases and risk factors stemming from, or associated with, endothelial dysfunction such as acute coronary syndromes (32), atherosclerosis (39), heart failure (16), stroke (44), hypertension (53), dyslipidemia (39), diabetes (48), and obesity (46). Moreover, elevations in EMVs have been shown to be indicative of poor outcome in a variety of settings, including heart transplant, myocardial infarction, and pulmonary hypertension (1, 11). Importantly, depending on their stimulus for release, EMVs can be pathogenic, inducing and perpetuating endothelial dysfunction resulting in the development of a proatherogenic prothrombotic vascular phenotype (43). For example, EMVs generated under diabetic glucose conditions have been shown to impair endothelial cell nitric oxide production and ignite cellular inflammation (7, 26). Currently it is unknown whether HIV-1 viral proteins stimulate EMV release and, if so, whether the viral protein-derived EMVs are deleterious to endothelial cell function.

Accordingly, we tested the hypotheses that: 1) HIV-1 Bal (R5)- and Lav (X4)-tropic gp120 and Tat stimulate the release of microvesicles from endothelial cells in vitro; and 2) viral protein-induced EMVs are deleterious to endothelial cell function, promoting endothelial cell inflammation, oxidative stress, senescence, and apoptosis. Understanding the role of EMVs may provide additional mechanistic insight into the adverse vascular effects of HIV-1-associated proteins gp120 and Tat.

METHODS

Viral proteins.

Recombinant HIV-1 R5 gp120 and Tat were obtained through the AIDS Research and Reference Reagent Program (Division of AIDS, National Institutes of Health), and HIV-1 X4 gp120 was acquired from Protein Sciences (Meriden, CT). To reconstitute Tat, 100 ml of PBS were bubbled with compressed nitrogen gas for 20 min followed by the addition of 15 mg of DTT and 100 mg of BSA and cooled on ice. Thereafter, Tat was dissolved in 250 μl of PBS solution. R5 and X4 were diluted in culture media to the desired concentrations.

Cell culture and EMV generation.

Human aortic endothelial cells (HAECs) (Life Technologies, ThermoFisher, Waltham, MA) were cultured in endothelial growth media (EBM-2 BulletKit) (Lonza, Basel, Switzerland) supplemented with 100 U/ml penicillin and 100 μg/ml streptomycin under standard cell culture conditions (37°C and 5% CO₂). Growth media was replaced 24 h after initial culture and every 2 days thereafter. Cells were serially passaged after reaching 80–90% confluence, and cells were harvested for experimentation after reaching ~90% confluence on the third or fourth passage. Cells were seeded in tissue culture flasks (Falcon, Corning, NY) and treated with media in the absence (control) and presence of HIV-1 R5 (100 ng/ml), X4 (100 ng/ml), or Tat (500 ng/ml) for 24 h (n = 4 experimental units). The control condition was used to generate viral protein-naïve EMVs. Viral protein concentrations were similar to circulating levels in untreated HIV-1-seropositive adults (36). After 24 h, media from each condition (control, R5, X4, and Tat) were collected, and EMV concentration was determined by flow cytometry. Remaining media were stored at −80°C for EMV isolation and EMV-related experiments. No further experimentation was performed on the viral protein-treated or control cells.

EMV characterization and enumeration.

Collected media from each condition was centrifuged at 13,000 g at room temperature for 2 min to pellet and discard cellular debris. Thereafter, 100 µl of the cell free supernatant were transferred to TruCount tubes (BD Biosciences), incubated with the fluorochrome-labeled antibody CD144-phycoerythrin (VEcadherin), fixed with paraformaldehyde (ChemCruz Biochemicals, Santa Cruz, CA), and diluted with 500 µl of PBS (4). All samples were analyzed using a FACS Aria I flow cytometer (BD Biosciences). EMV size threshold (0.16–1 µM diameter) was established using Megamix-Plus SSC calibrator beads (Biocytex, Marseille, France). EMVs from R5-, X4-, and Tat-treated cells were determined using the formula: (number of events in region containing EMVs/number of events in absolute count bead region) × (total number of beads per test/total volume of sample).

EMV-treated cells.

HAECs were cultured as described above and seeded in six-well tissue culture plates. Media containing EMVs collected from the control condition and the R5-, X4-, and Tat-stimulated cells were centrifuged at 13,000 g at room temperature for 2 min to pellet and discard cellular debris. Thereafter, the supernatant was centrifuged at 20,500 g for 30 min at 4°C to pellet EMVs (4, 26). EMVs were resuspended in media at a concentration of 1.0 × 10⁷ EMV/ml. HAECs were treated with media containing EMVs from the control, R5, X4, or Tat conditions for 24 h. Cells were treated with EMVs in a 1:2 cell-to-microvesicle ratio. The final concentration of EMVs in the media was equal between the conditions and ranged from 200 to 1,200 EMVs/µl, dependent on the number cells used for each experimental condition. After treatment, cells and media were harvested for the determination of cell inflammation, oxidative stress, senescence, apoptosis, and intracellular micro-RNA (miR) expression as well as cytokine release.

Cellular inflammation.

Concentration of tumor necrosis factor-α (TNF-α), interleukin (IL)-6, and IL-1β was quantified in the media harvested from each condition using a chemiluminescent ELISA (R&D Systems, Minneapolis, MN) (22). Cell surface expression of intracellular adhesion molecule-1 (ICAM-1) was determined by flow cytometry (20). Intra-assay coefficient of variation for the chemiluminescent ELISAs was <8% for each assay.

Intracellular oxidative stress.

HAECs were seeded in 96-well tissue culture plates (Thermo Scientific) and allowed to adhere overnight. Adherent cells were washed and treated with 2′,7′-dichlorofluorescin diacetate (DCFDA) (Abcam, Cambridge, MA) stain (25 μM) for 45 min. After DCFDA treatment, cells were washed two times and stimulated with cell culture media or media containing X4-EMVs, R5-EMVs, or Tat-EMVs for 3 h. Immediately thereafter, fluorescence was measured using a GEMINI EM microplate reader (Molecular Devices, Sunnyvale, CA) and reported as mean fluorescence intensity (MFI) (54).

Senescence-associated β-galactacidase assay.

Cellular senescence was quantified using cytochemical senescence-associated β-galactacidase (SA-β-gal) staining. Following EMV treatment, HAECs were washed two times with PBS and incubated with 2 ml of fixative (2% formaldehyde and 0.2% glutaraldehyde) for 5 min. Fixed cells were washed two times with PBS and then incubated for 14 h with 2 ml of freshly prepared staining solution (1 mg/ml 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside in dimethylformamide, 40 mM citric acid/sodium phosphate, 5 mM potassium ferrocyanide, 5 mM potassium ferricyanide, and 150 mM NaCl, 2 mM MgCl₂; ThermoFisher). The staining solution was then removed, and cells were washed two times with PBS and one time with methanol and allowed to air-dry. Cells were visualized by light microscopy (Zeiss, Thornwood, NY) and quantified in five random image fields for each condition. Cells with blue cytoplasmic staining were identified as senescent-positive cells. Senescent cells (%) were determined as SA-β-gal-positive cells divided by the total number of cells counted (13).

Intracellular active caspase-3.

After EMV treatment, harvested cells (3.0 × 10⁶) for caspase-3 determination were treated with staurosporine (1 µmol/l) for 3 h at 37°C and biotin-ZVKD-fmk inhibitor for 1 h at 37°C. Intracellular concentration of active caspase-3 was determined by enzyme immunoassay (31).

Intracellular miR expression.

Intracellular miR expression was determined by RT-PCR. After EMV treatment, 1.0 × 10⁵ cells were harvested, and total cellular RNA was isolated using the miRVANA RNA isolation kit (Exiqon, Vedbake, Denmark) (21). RNA concentration was determined using a Nanodrop Lite spectrophotometer (ThermoFisher). Thereafter, 150 ng of RNA were reverse transcribed using the miScript II Reverse Transcription Kit (Qiagen, Hilden, German) (21). RT-PCR was performed using the Bio-Rad CFX96 RT-PCR platform with the miScript SYBR green PCR kit (Qiagen) and specific primers miR-34a, miR-126, miR-146a, miR-181b, miR-221, miR-Let-7a, and U6 (Qiagen) (21). All samples were assayed in duplicate. miR expression was quantified using the comparative C_t method and normalized to U6. The fold change of each transcript was calculated as the ${{\displaystyle 2}}^{-\Delta \Delta {\text{C}}_{\text{t}}}$ where fold change (arbitrary units) =  ${{\displaystyle 2}}^{{}^{-\left[{\text{C}}_{\text{t}}\left(\text{miR\hspace{0.17em}experimental}\right)-{\text{C}}_{\text{t}}\left(\text{RNU6\hspace{0.17em}experimental}\right)-{\text{C}}_{\text{t}}\left(\text{miR\hspace{0.17em}contol}\right)-{\text{\hspace{0.17em}C}}_{\text{t}}\left(\text{RNU6\hspace{0.17em}control}\right)\right]}}$.

Statistical analysis.

Differences between treatments were determined by analysis of variance. Where indicated by a significant F-value, Student-Newman-Keuls post hoc tests were performed. Data are reported as means ± SE for four independent HAEC experiments. Statistical significance was set a priori at P < 0.05.

RESULTS

EMV release.

EMV release in response to each viral protein is shown in Fig. 1. The number of EMVs generated from HAECs treated with R5 (141 ± 21 MV/µl), X4 (132 ± 20 MV/µl), and Tat (130 ± 20 MV/µl) was significantly higher (~120%, ~130%, and ~120%, respectively) compared with EMVs released under control culture conditions (control: 61 ± 13 MV/µl). There were no significant differences in EMV generation among the viral proteins.

Fig. 1.

Effect of human immunodeficiency virus (HIV)-1 R5 glycoprotein (gp) 120, X4 gp120, and transactivator of transcription (Tat) on endothelial cell microvesicle release. EMV, endothelial microvesicle. Values are means ± SE (n = 4). *P < 0.05 vs. control (untreated) culture conditions.

Endothelial inflammation.

EMVs from the viral proteins induced higher endothelial cell release of proinflammatory cytokines and greater upregulation in the surface expression of ICAM-1 than control (Fig. 2). TNF-α release was significantly higher (~25–30%) in response to R5-EMVs (5.75 ± 0.14 pg/ml), X4-EMVs (5.91 ± 0.09 pg/ml), and Tat-EMVs (5.79 ±0.42 pg/ml) compared with control EMVs (3.94 ±0.18 pg/ml; Fig. 2A). Similarly, release of IL-6 and IL-1β (Fig. 2, B and C) was significantly greater (~80% and ~25%, respectively) from cells treated with R5-EMVs (103.66 ± 4.15 and 2.56 ± 0.12 pg/ml, respectively), X4-EMVs (94.11 ±6.60 and 2.72 ± 0.18 pg/ml), and Tat-EMVs (94.55 ± 11.34 and 2.59 ± 0.06 pg/ml) than cells treated with control EMVs (56.26 ± 13.53 and 2.11 ± 0.14 pg/ml). Cell surface expression of ICAM-1 was approximately twofold higher (P < 0.05) in HAECs treated with R5-EMVs (2,808.0 ± 413.1 MFI), X4-EMVs (3,016.0 ± 343.1 MFI), and Tat-EMVs (2,868.8 ±479.1 MFI) compared with control EMVs (1,512.1 ± 135.2 MFI; Fig. 2E). Of note, the magnitude of increase in TNF-α, IL-6, and IL-1β release and ICAM-1 expression was not significantly different between the viral protein EMV treatment conditions.

Fig. 2.

Effect of endothelial microvesicles (EMVs) induced by human immunodeficiency virus (HIV)-1 R5 glycoprotein (gp) 120, X4 gp120, and transactivator of transcription (Tat) on endothelial cell release of tumor necrosis factor-α (TNF-α, A), interleukin (IL)-6 (B), and IL-1β (C). Representative dot plots and flow cytometric histograms (D) used to quantify cell surface expression of intracellular adhesion molecule-1 (ICAM-1, E). Values are mean ± SE (n = 4). *P < 0.05 vs. EMVs from control culture condition.

Endothelial oxidative stress.

Intracellular reactive oxygen species production was higher (~45%; P < 0.05) in the R5-EMV (76.7 ± 3.8 MFI)-, X4-EMV (82.8 ± 4.2 MFI)-, and Tat-EMV (72.9 ± 1.4 MFI)- compared with control EMV (54.2 ± 3.4 MFI)-treated cells (Fig. 3). The production of reactive oxygen species was similar between endothelial cells treated with R5-EMVs, X4-EMVs, and Tat-EMVs.

Fig. 3.

Effect of endothelial microvesicles (EMVs) induced by human immunodeficiency virus (HIV)-1 R5 glycoprotein (gp) 120, X4 gp120, and transactivator of transcription (Tat) on endothelial cell reactive oxygen species production. ROS, reactive oxygen species. Values are means ± SE (n = 4). *P < 0.05 vs. EMVs from control culture condition.

Endothelial senescence.

The percentage of SA-β-gal-stained cells was markedly higher (~50–70%; P < 0.05) in cells treated with R5-EMVs (36.1 ± 2.2%), X4-EMVs (32.8 ±2.9%), and Tat-EMVs (31.9 ± 2.4%) compared with control EMVs (22.3 ± 2.4%) (Fig. 4). Endothelial cell senescence was not significantly different between the cells treated with the viral protein-induced EMVs.

Fig. 4.

Representative images of senescence-associated β-galactacidase (SA-β-gal)-stained control endothelial microvesicle (EMVs) and Bal glycoprotein 120 (R5)-EMV-treated cells. Cellular senescence was quantified in all conditions by a single blinded investigator. Effect of EMVs induced by human immunodeficiency virus (HIV)-1 R5, X4 glycoprotein 120, and transactivator of transcription (Tat) on endothelial cell senescence. Values are means ± SE (n = 4). *P < 0.05 vs. EMVs from control culture condition.

Endothelial apoptosis.

Staurosporine-stimulated intracellular active caspase-3 levels were significantly higher (~70%) in R5-EMV (1.61 ± 0.15 pg/ml)-, X4-EMV (1.53 ± 0.21 pg/ml)-, and Tat-EMV (1.59 ± 0.22 pg/ml)-treated cells compared with control EMV (0.93 ± 0.24 pg/ml)-treated cells (Fig. 5). There was no significant difference in intracellular active caspase-3 concentrations between the viral protein EMV-treated HAECs.

Fig. 5.

Effect of endothelial microvesicles (EMVs) induced by human immunodeficiency virus (HIV)-1 R5 glycoprotein (gp) 120, X4 gp120, and transactivator of transcription (Tat) on endothelial cell staurosporine-stimulated active caspase-3. Values are means ± SE (n = 4). *P < 0.05 vs. EMVs from control culture condition.

Endothelial miR expression.

Table 1 presents endothelial cell expression of miR-34a, miR-126, miR-146a, miR-181b, miR-221, and miR-Let-7a in the R5-EMV-, X4-EMV-, and Tat-EMV-treated cells. Endothelial expression of miR-34a was significantly higher (~50%) in cells treated with R5-, X4-, and Tat-EMVs compared with control EMVs. Cellular expression of miR-126, miR-146a, miR-221, and miR-Let-7a was significantly reduced (~35–120%) in response to R5-EMV, X4-EMV, and Tat-EMV. miR-181b was lower in R5-EMV- and Tat-EMV-treated cells; however, there was no significant effect of X4-EMV on the cellular expression of miR-181b.

Table 1.

Endothelial cell miR expression

miR	R5-EMVs (fold vs. control)	X4-EMVs (fold vs. control)	Tat-EMVs (fold vs. control)
34a	1.55 ± 0.15*	1.48 ± 0.12*	1.52 ± 0.08*
126	−1.44 ± 0.06*	−1.79 ± 0.08*	−1.39 ± 0.10*
146a	−2.17 ± 0.10*	−1.52 ± 0.10*	−1.64 ± 0.09*
181b	−2.07 ± 0.15*	1.19 ± 0.06	−2.05 ± 0.14*
221	−1.97 ± 0.04*	−1.31 ± 0.05*	−1.67 ± 0.12*
Let-7a	−2.46 ± 0.11*	−2.19 ± 0.14*	−1.97 ± 0.10*

Values are means ± SE. miR, microRNA; R5, Bal glycoprotein 120; EMVs, endothelial microvesicles; X4, Lav glycoprotein 120; Tat, transactivator of transcription.

^*P < 0.05 vs. control EMV-treated cells.

DISCUSSION

The primary new findings of the present study are as follows: 1) HIV-1 R5, X4 gp120, and Tat induce microvesicle release from endothelial cells; and 2) R5-, X4-, and Tat-induced EMVs have deleterious effects on endothelial cell function. Indeed, viral protein-derived EMVs promote a proinflammatory, pro-oxidative, prosenescent, and proapoptotic endothelial phenotype. To our knowledge, this is the first study to determine the effects of HIV-1-associated proteins on microvesicle release from endothelial cells and the impact of viral protein-induced EMVs on endothelial cell function. EMVs released in response to gp120 and Tat provide novel insight into potential mechanisms contributing to the increased risk and prevalence of endothelial dysfunction and, in turn, atherosclerotic vascular disease with HIV-1 (33).

Indication that EMVs may be causative agents in vascular disease initiation and progression stem from their numerical increase in almost all vascular-related clinical and subclinical conditions as well as their functional properties under pathological conditions (5, 11). The release of EMVs is a result of plasma membrane budding due to disruption of membrane phospholipid asymmetry, cytoskeleton protein destabilization, and loss of functional integrity (12). Under normal healthy conditions, endothelial cells will release low levels of microvesicles to help maintain cell homeostasis and aid in cell-to-cell communication (19). However, under conditions of endothelial activation, injury, or apoptosis, EMV vesiculation and release are upregulated, and in most cases the functional phenotype of the EMV is dictated by the pathological nature of the triggering stimulus (10, 41). Clinical and experimental studies have demonstrated that, under pathological conditions, EMV release is increased, and the released EMVs can engage in a vicious cycle amplifying and perpetuating the deleterious vascular effects of the primary stimulus (6, 26, 38). In the present study, we demonstrate, for the first time, that HIV-1 R5, X4 gp120, and Tat significantly increase EMV release from cultured endothelial cells. Interestingly, the magnitude of increase in EMVs in response to each viral protein was almost identical (~135%), suggesting common release mechanisms. Indeed, both gp120 and Tat have been shown to activate various signaling proteins involved in EMV formation and release. For example, gp120 binding of CCR5 and/or CXCR4 (30) results in the activation of p38 mitogen-activated protein kinase (MAPK) and caspase-2 (18, 40), key initiators of EMV formation and release under conditions of cell activation and increased apoptotic susceptibility (27, 41). p38 MAPK activation and caspase-2 provoke cytoskeletal rearrangement, stimulate microvesicle vesiculation, and initiate release (41). In addition, Tat has also been shown to activate p38 MAPK and caspase-2 through either binding with cell surface receptors or interaction with cytosolic signaling proteins (23, 33, 42).

Endothelial activation and inflammation are harbingers of endothelial dysfunction and pivotal to the development of atherosclerosis (28, 29). Inflammation mediators such as cell adhesion molecules and cytokines promote the recruitment, attachment, and translocation of inflammatory cells in the subendothelial space and the subsequent retention of lipids, thereby initiating atherosclerotic lesion development (29). Upregulation of cell adhesion molecules on the surface of the endothelium is an early and sensitive indicator of endothelial cell activation, distress, and inflammation (28). Herein, we demonstrate that R5-, X4- and Tat-induced EMVs significantly increased the surface expression of ICAM-1 (~30%) on treated HAECs. Increased expression of ICAM-1 on the surface of endothelial cells is known to facilitate greater luminal leukocyte interaction, adhesion, and migration in the subendothelial space, increasing the propensity for atherosclerosis (28). In addition to increasing ICAM-1 expression on the endothelial cell surface, each of the viral protein-derived EMVs also induced significant endothelial release of proinflammatory cytokines. Concentrations of TNF-α (~25%), IL-6 (~80%), and IL-1β (~25%) were markedly higher in the media of cells treated with R5-, X4-, and Tat-EMVs. Increased production and release of TNF-α, IL-6, and IL-1β by the endothelium has been linked with the initiation and development of endothelial dysfunction and, ultimately, atherosclerotic lesion formation (28, 29). Interestingly, the observed inflammatory effects of viral protein-induced EMVs on endothelial cells are similar to the endothelial effects of the viral proteins themselves and may contribute to the reported elevations in circulating levels of TNF-α, IL-6, IL-1β, and soluble ICAM-1 in HIV-1-seropositive adults (3, 52).

Increased oxidative stress often accompanies a hyperinflammatory state (24). As such, it is not surprising that endothelial ROS production was markedly increased (~45%) in cells treated with either R5-, X4-, or Tat-EMVs. Excess ROS production can damage or modify intracellular proteins and react with nitric oxide to form peroxynitrite, impairing cell function and viability (24). In fact, oxidative stress has been implicated as a key mechanism underlying viral protein-mediated endothelial dysfunction (14, 40). This deleterious effect of gp120 and Tat on the endothelium may also be instigated by the microvesicles they generate.

Increased cellular senescence and apoptotic susceptibility often mark the penultimate stage in declining cell vitality, function, and survival (15, 34). The initiation of senescent and/or apoptotic processes and pathways can lead to pathological circumstances, depending on the stimulus. In the case of endothelial cells, senescence and apoptosis result in the arrest and/or termination of key atheroprotective functions rendering the endothelium prone to atherosclerosis and thrombosis (15, 34). Both human and animal studies have demonstrated that endothelial cell senescence is remarkably higher at sites of coronary and aortic atherosclerosis compared with nondiseased areas of the vessel (35, 51). Indeed, a high preponderance of senescent endothelial cells has been reported in vivo at atherosclerotic sites in both the aorta and coronary arteries (35, 51). We (21, 31) and others (50) have previously demonstrated that gp120 and Tat induce both a prosenescent and proapoptotic endothelial phenotype. The results of the present study compliment and significantly extend these findings by demonstrating that EMVs generated by each of these viral proteins also provoke endothelial cell senescence and heightened apoptotic susceptibility. The percentage of SA-β-gal-stained HAECs was substantially higher in the cells exposed to R5-, X4-, and Tat-induced EMVs. Concomitant with the increase in cell senescence, the viral protein-related EMVs augmented endothelial cell apoptotic susceptibility. In response to the apoptotic stimulus, staurosporine, intracellular active caspase-3 concentrations were significantly higher in the cells incubated with each viral protein-stimulated EMV compared with control condition EMVs. Concentrations of active caspase-3, the so called “executioner molecule” in the hierarchy of caspases, provide specific and sensitive indication of the apoptotic tendency of a cell (50). Interestingly, the degree of cell senescence and active caspase 3 noted in the present study is comparable to the levels we have previously reported in response to R5, X4, and Tat (21, 50), providing further evidence that EMVs generated by these viral proteins are as toxic to the endothelium as the direct endothelial effects of the proteins per se.

The mechanisms underlying the profound deleterious effects of the viral protein-derived EMVs on endothelial cell function are unknown. To identify potential mechanisms and possible pathways underlying the proinflammatory, pro-oxidative, prosenescent, and proapoptotic effect of R5-EMVs, X4-EMVs, and Tat-EMVs, we examined changes in endothelial cell miR profiles. miRs play a central role in regulating vascular health and function through the posttranscriptional regulation of gene expression (47). Altered endothelial cell expression of vascular- and inflammation-related miRs, specifically miR-34a, miR-126, miR-146a, miR-181b, miR-221, and miR-Let-7a, has been shown to contribute mechanistically to endothelial inflammation, oxidative stress, senescence, and apoptosis (19, 37, 47). In regard to viral protein EMV-mediated changes in cell inflammation, the expression of miR-146a and miR-221 was markedly disrupted by each of the viral protein EMVs. Both miR-146a and miR-221 expression was significantly lower in the R5-, X4-, and Tat-EMV-treated cells. The reduction in miR-146a across all of the viral protein EMV conditions is particularly striking because miR-146a directly targets activators of nuclear factor-κB (NF-κB), such as tumor necrosis factor receptor-associated factor-6 and interleukin-1 receptor-associated kinase-1 (9), thereby suppressing its activation. NF-κB is a primary proinflammatory transcription factor regulating cytokines such as TNF-α, IL-6, and IL-1β as well as the cellular expression of ICAM-1 (9). Thus, reduction in miR-146a expression may be an underlying factor contributing to the observed increase in soluble cytokine release and ICAM-1 expression observed in the R5-, X4-, and Tat-EMV-treated cells. Interestingly, miR-181b expression was diminished (~100%) in response to EMVs induced by R5 gp120 and Tat but not X4 gp120. miR-181b directly targets the NF-κB transporter importin-α₃, suppressing NF-κB translocation from the cytoplasm to the nucleus and, in turn, NF-κB-mediated transcriptional activity (47). Reasons for the disparate effect of X4-EMVs on cellular miR-181b expression compared with R5- and Tat-EMVs are not readily apparent. However, disparities in the vascular effects of R5 and X4 are not uncommon and have been noted in other experimental conditions (30). In addition to reductions in the aforementioned miRs, expression of miR-34a was higher, whereas miR-126 and miR-Let-7a expression was lower in the cells treated with each of the viral protein-derived EMVs. Elevation in miR-34a expression promotes cellular senescence and apoptosis through the negative regulation of sirtuin-1 and the anti-apoptotic protein BCL-2, respectively (37). Lower miR-126 and miR-Let-7a expression further fosters a proapoptotic cellular phenotype through dysregulation of caspase-3 activity (8, 49). Clearly, future studies are required to more carefully dissect the mechanisms responsible for the deleterious effects of gp120- and Tat-induced EMVs. The observed disruption in miR expression provides valuable clues and potential cell signaling pathways and proteins to target.

There is one experimental limitation regarding the present study that deserves to be mentioned. Unfortunately, we were unable to determine the combined effects of the viral proteins on EMV release and phenotype. The experiments performed in the present study exhausted our supply of R5 and Tat, which is capped annually by the National Institutes of Health AIDS Research and Reference Reagent Program. However, considering EMV release is dependent on activation- or apoptosis-related cellular pathways, it is unclear if additional stimuli would induce a synergistic effect. There are some data to suggest an all-or-none phenomena for EMV release (2). Thus, it is plausible that EMV release to the viral proteins would not differ with combined treatment. We would, however, expect the EMV phenotype to be altered by the stimulus for release. Testing the effect of combined viral proteins remains an important future direction.

In conclusion, the results of this study demonstrate that HIV-1 gp120 and Tat proteins induce microvesicle release from endothelial cells, and these microvesicles confer pathological effects on endothelial cells by inducing inflammation, oxidative stress, and senescence as well as enhancing susceptibility to apoptosis. It is well established that these phenotypic changes in endothelial cells are associated with, and mediate, the development of atherosclerosis and thrombosis (17, 35, 41). The noted independent effects of HIV-1 R5 and X4 gp120 as well as Tat on EMVs provide novel additional insight regarding the atherogenic properties of these endothelial toxic viral proteins.

GRANTS

This study was supported, in part, by National Heart, Lung, and Blood Institute Grant HL-131458.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.G.H., K.A.S., M.V.L., L.M.B., T.D.B., and J.J.G. performed experiments; J.G.H., K.A.S., M.V.L., L.M.B., J.J.G., and C.A.D. analyzed data; J.G.H., T.D.B., J.J.G., and C.A.D. interpreted results of experiments; J.G.H. and J.J.G. prepared figures; J.G.H. and C.A.D. drafted manuscript; J.G.H., J.J.G., E.C., and C.A.D. edited and revised manuscript; J.G.H., K.A.S., M.V.L., L.M.B., T.D.B., J.J.G., E.C., and C.A.D. approved final version of manuscript; C.A.D. conceived and designed research.