Novel Mechanism of Microvesicle Regulation by the Antiviral Protein Tetherin During HIV Infection

Abstract

Background

Microvesicles are cell membrane–derived vesicles that have been shown to augment inflammation. Specifically, monocyte‐derived microvesicles (MDMVs), which can express the coagulation protein tissue factor, contribute to thrombus formation and cardiovascular disease. People living with HIV experience higher prevalence of cardiovascular disease and also exhibit increased levels of plasma microvesicles. The process of microvesicle release has striking similarity to budding of enveloped viruses. The surface protein tetherin inhibits viral budding by physically tethering budding virus particles to cells. Hence, we investigated the role of tetherin in regulating the release of MDMVs during HIV infection.

Methods and Results

The plasma of aviremic HIV‐infected individuals had increased levels of tissue factor + MDMVs, as measured by flow cytometry, and correlated to reduced tetherin expression on monocytes. Superresolution confocal and electron microscopy showed that tetherin localized at the site of budding MDMVs. Mechanistic studies revealed that the exposure of monocytes to HIV‐encoded Tat triggered tetherin loss and subsequent rise in MDMV production. Overexpression of tetherin in monocytes led to morphologic changes in the pseudopodia directly underneath the MDMVs. Further, tetherin knockout mice demonstrated a higher number of circulating MDMVs and less time to bleeding cessation.

Conclusions

Our studies define a novel regulatory mechanism of MDMV release through tetherin and explore its contribution to the procoagulatory state that is frequently observed in people with HIV. Such insights could lead to improved therapies for individuals infected with HIV and also for those with cardiovascular disease.

Nonstandard Abbreviations and Acronyms

cART

combined antiretroviral treatment

INSTI

integrase strand transfer inhibitor

MDMVs

monocyte‐derived microvesicles

MFI

median fluorescence intensity

NNRTI

nonnucleoside reverse transcriptase inhibitor

NRTI

nucleoside reverse transcriptase inhibitor

PI

protease inhibitor

TF

tissue factor

Clinical Perspective

What Is New?

• Our data demonstrate a novel relationship between the surface protein and the release of tissue factor–positive microvesicles from monocytes during HIV infection.

• Treatment of monocytes with HIV Tat recapitulated what was observed in clinical monocytes from those with HIV: (1) increased tissue factor expression, (2) decreased tetherin expression, and (3) increased release of microvesicles.

What Are the Clinical Implications?

• Future therapies could target tetherin to modulate microvesicle release, providing a novel treatment to ameliorate inflammatory aspects relating to HIV comorbidities and potentially for other microvesicle‐related inflammatory diseases as well.

Age‐related inflammatory diseases are dramatically increased in the HIV‐infected population despite the success of combined antiretroviral treatment (cART) in repressing viral loads. ¹ , ² , ³ One major contributor to this elevation is the activation of monocytes in patients with aviremic HIV, as evidenced by elevated levels of soluble monocyte activation markers sCD163 ⁴ and sCD14 ⁵ in plasma. Furthermore, both sCD163 and sCD14 have been linked to increased incidence of cardiovascular disease, ⁶ , ⁷ neurocognitive disorders, ⁸ , ⁹ and mortality ¹⁰ , ¹¹ , ¹² in HIV‐infected individuals. One insufficiently studied mechanism of monocyte‐augmented inflammation is the release of microvesicles.

Microvesicles, small membrane–derived extracellular vesicles ranging from 100 to 1000 nm in diameter, detach from the plasma membrane at lipid rafts ¹³ , ¹⁴ , ¹⁵ and can contain surface receptors, ¹⁶ proteins, ¹⁷ lipids, ¹⁸ mRNA, ¹⁹ and mitochondria ²⁰ derived from the cell of origin. Acting as signaling shuttles, microvesicles communicate with recipient cells by binding to surface receptors or by delivering cargo through endocytosis. ²¹ Microvesicles promote inflammation through multiple mechanisms, such as activating endothelial cells ²² , ²³ or augmenting coagulation through tissue factor (TF) expression. ²⁴ , ²⁵ Higher rates of circulating microvesicles are linked to numerous inflammatory diseases, ²⁶ , ²⁷ , ²⁸ , ²⁹ , ³⁰ , ³¹ many of which are prevalent in the HIV‐infected population. ¹ , ² , ³ , ³² Studies have also demonstrated that microvesicles from treated HIV‐infected individuals increase inflammation in endothelial cells. ³³ Despite their important impact during inflammation, little is understood about the processes regulating microvesicle biogenesis and release, particularly those emerging from monocytes.

The plasma membrane protein tetherin (also referred to as BST2, CD317, and HM1.24) regulates virion release ³⁴ and provides a potential unstudied mechanism of microvesicle regulation. Using its dual‐membrane domains, tetherin physically sequesters budding enveloped viruses to the cell surface. As virions bleb from the plasma membrane at lipid rafts, ³⁵ either of tetherin's terminals are inserted in the envelope of the budding virion, thereby restricting viral release until both tetherin and the virion are recycled back into the cell for degradation. ³⁶ , ³⁷ Tetherin may additionally tether microvesicles, as their egress from the cell has striking similarities to virion release, as both occur at the curvature formation of lipid rafts. ¹³ , ³⁵ A recently published study demonstrated that overexpression of tetherin in cultured cell lines can also tether nanoparticles (also known as exosomes) that are released when endocytic compartments fuse with the plasma membrane. ³⁸ Although microparticles and nanoparticles are similar in their ability to act as signaling shuttles and induce inflammation, these particles greatly differ in their biogenesis from the secreting cell. Microparticles arise from the plasma membrane, whereas nanoparticles emerge from multivesicular bodies. ³⁹ Thus, the focus of this study is to investigate tetherin's effect on microvesicle regulation.

In this report, we investigated tetherin's ability to sequester monocyte‐derived microvesicles (MDMVs) to the surface of the cell during HIV infection. We demonstrated that tetherin levels are decreased on the surface of CD14+ monocytes in HIV‐infected individuals receiving cART and that these levels negatively correlate with the amount of circulating MDMVs in the plasma. In vitro studies performed with primary monocytes established that treatment with the HIV‐encoded protein Tat causes decreased tetherin expression on the surface of these cells and increased MDMV release. Further, a direct relationship between MDMVs and tetherin was established when mice lacking tetherin had elevated levels of circulating MDMVs and exhibited reduced time of bleeding cessation. This study investigates a novel mechanism of MDMV regulation by tetherin, discovering new targets for improved treatments in those with HIV and other microvesicle‐related inflammatory diseases.

Methods

Data that support the findings of this study are available from the corresponding author upon reasonable request.

Ethics Statement

All studies involving human samples were reviewed and approved by the University of Rochester Medical Center Research Subjects Review Board. Blood samples were acquired from adults after written informed consent in accordance with the Declaration of Helsinki. Animal studies were performed in accordance with the Animal Welfare Act and the National Institute of Health guidelines, and protocols were reviewed and approved by the University Committee on Animal Resources of the University of Rochester Medical Center. The Association for the Assessment and Accreditation of Laboratory Animal Care International fully accredited the facilities and programs of the Vivarium and Division of Laboratory Animal Medicine of the School of Medicine and Dentistry. The data that support the findings of this study are available from the corresponding author.

Whole Blood Samples From Human Subjects

Participants with or without HIV infection were eligible for the study if they had not previously taken any nonsteroidal anti‐inflammatory drugs within a week of blood donation. Blood was drawn into acid citrate dextrose buffered vacutainers (BD Biosciences, San Jose, CA). Participant demographics are reported in Table.

Table 1.

Donor Demographics and Characteristics

Characteristics	HIV−	HIV+
Total number of participants	20	20
Mean age, y (SD)	47 (11)	53 (9)
Sex, N (%)
Female	12 (60)	3 (15)
Male	8 (40)	17 (85)
Race, N (%)
Black	5 (25)	6 (30)
Hispanic	0 (0)	2 (10)
White	15 (75)	12 (60)
Mean CD4 count, cells/mm3 (SD)		859 (239)
cART treatment, N (%)
NRTI/PI		7 (35)
NRTI/NNRTI		6 (30)
NRTI/INSTI		6 (30)
PI/NNRTI		1 (5)
Viral load
Undetectable, N (%)		17 (85)
Detectable, N (%, mean RNA copies/mL [SD])		3 (15%, 80 [53])

cART indicates combined antiretroviral therapy; INSTI, integrase strand transfer inhibitor; NNRTI, nonnucleoside reverse transcriptase inhibitor; NRTI, nucleoside reverse transcriptase inhibitor; and PI, protease inhibitor.

Flow Cytometry Analysis

Monocytes were analyzed as previously described. ⁴⁰ Whole blood was fixed with paraformaldehyde (2% final concentration) followed by lysis of red blood cells using ACK RBC lysis buffer (Gibco, Los Gatos, CA). Samples were then stained with 3 μL anti‐CD16 PE Cy7 (BD Biosciences, Franklin Lakes, NJ), 8 μL anti‐CD14 PE (BD Biosciences), and 5 μL anti‐CD317 Alexa Fluor 488 (Tetherin, Invitrogen, Grand Island, NY) for 30 minutes following acquisition using an Accuri C6 flow cytometer (BD Biosciences). In vitro flow cytometry was performed with monocytes fixed as above and stained with 5 μL anti‐CD317 Alexa Fluor 488 (Tetherin, Invitrogen) or 5 μL anti‐CD142 APC (TF, Invitrogen). Fluorescence minus one controls were used to define gates.

Microvesicles were quantified from platelet‐free plasma. Whole blood was loaded onto Histopaque (Sigma Aldrich, Saint Louis, MO) and centrifuged at 1250g  for 25 minutes. Platelet‐poor plasma was collected and centrifuged at 5000g for 5 minutes to collect platelet‐free plasma. The platelet‐free plasma diluted in 0.2 μm filtered Dulbecco's phosphate buffered saline (Sigma Aldrich, St Louis, MO) and stained with 8 μL anti‐CD14 PE, 5 μL anti‐CD14 FITC (Invitrogen) and 5 μL anti‐CD147 APC (TF, Invitrogen) for 30 minutes. Twenty‐five μL was acquired using the Accuri C6 cytometer, and sizing beads (Mega Mix, BioCytex, Marseille, France) ranging from 500 to 1000 nm were used to define the gate for microvesicles.

The mouse strain C57BL/6 was purchased from the Jackson Laboratory. The tetherin‐knockout mouse strain B6.Cg‐BST2^tm1.Bsz/J was a gift from Dr Paul Biendiaz (Rockefeller University) and was generated as previously described. ⁴¹ For quantification of microvesicles in mouse plasma, blood was obtained by retro‐orbital puncture. platelet‐free plasma was collected by centrifugation at 3500g for 5 minutes and stained using 0.2 μL anti‐CD115 APC (eBioscience, San Diego, CA) per sample.

In Vitro Treatment of Monocytes and MDMV Quantification

Monocytes were isolated from HIV‐negative donors by MACS CD14+ Beads (Miltenyi Biotec, Sunnyvale, CA) per manufacturer's instructions and cultured with 0.2 μm filtered RPMI 1640 supplemented with 10% fetal bovine serum and 1% penicillin‐streptomycin‐glutamine. When specified, cells were treated with 500 nmol/L HIV viral full‐length recombinant 101 amino acid Tat (kindly provided by the University of Rochester Center for AIDS Research), a dose used previously in the literature in vitro. ⁴² , ⁴³ , ⁴⁴ The National Institutes of Health AIDS Reagent Program provided HIV treatment drugs raltegravir (5 μmol/L), tenofovir disoproxil fumarate (5 μmol/L), and emtricitabine (5 μmol/L). Dimethyl sulfoxide (Sigma Aldrich) was used as vehicle control. After 18 hours, monocytes were processed for flow cytometry as described above. For Tat treatments, cells were stained for surface tetherin or TF as described above. For HIV drug treatments, cells were stained with 10 μL anti‐CD142‐FITC (TF, Miltenyi Biotec, Bergisch Gladbach, Germany).

Before quantification, cells and debris were depleted from the media by centrifugation. Where specified, vesicles were quantified using a Nanosight NS300 (Malvern Instruments, Amesbury, UK) and subsequently analyzed by NanoSight Tracking Analysis (Malvern Instruments). Briefly, media were diluted 1:10 in 0.2 μm filtered Dulbecco's phosphate buffered saline, and 5 videos lasting 30 seconds in duration were acquired. The number of background vesicles detected in the filtered media were subtracted for each experiment. Alternatively, media were stained with FilmTracer calcein green biofilm stain (Molecular Probes, Eugene, OR) according to manufacturer's specifications or 5 μL Annexin V APC (BD Biosciences) and 8 μL CD14 PE. Before Annexin V staining, microvesicles and RPMI was spun at 20 000g for 30 minutes and resuspended in 100 μL Annexin V binding buffer (BD Biosciences). Fifteen microliters of media was acquired on the BD Accuri C6 as described above.

mRNA Quantification

RNA from either Tat treated or nontreated cells was harvested using Trizol (Invitrogen) per the manufacturer's instructions, and contaminating DNA was digested using DNase I (Invitrogen). The iScript cDNA Synthesis Kit (Bio‐Rad, Hercules, CA) was used to synthesize RNA into cDNA as per the manufacturer's instructions. Power SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA) along with previously published primers for tetherin ⁴⁵ or the housekeeping gene GAPDH ⁴⁶ were analyzed on CFX Connect Real Time System (Bio‐Rad).

Lentivirus Transduction of Primary Monocytes

Expression plasmids pCDH‐CMV‐MCS‐EF1 PURO (System Biosciences, Palo Alto, CA) containing full‐length wild‐type tetherin transcripts were kindly provided by Dr Jean Gustin. ⁴⁷ , ⁴⁸ The tetherin expression plasmids, along with the packing plasmid pCMV4‐Δ89.2 and the envelope plasmid VSV‐G (vesicular stomatitis virus glycoprotein), were used to produce recombinant lentivirus vectors in HEK 293T cells. These cells were cultured in Dulbecco's Modified Eagle Medium supplemented with 10% fetal bovine serum and 1% penicillin‐steptomycin‐glutamine. Viruses were then concentrated using PEG6000 (Sigma Aldrich), ⁴⁹ and titered by using optical density as outlined. ⁵⁰ Finally, the primary cultures of human monocytes were transduced as recently described. ⁵¹

Electron Microscopy Imaging

For scanning electron microscopy, monocytes were isolated as described above and plated onto glass poly‐l‐lysine‐coated coverslips. Monocytes were treated with 20 ng/mL tumor necrosis factor alpha (Sigma) for 18 hours to induce microvesicle formation. Cells for immunogold labeling were then fixed with 1% paraformaldehyde for 30 minutes followed by blocking in 1% bovine serum albumin and 1% goat serum in Dulbecco's phosphate buffered saline for 30 minutes. Cells were incubated in a 1:500 dilution of goat anti‐tetherin antibody (National Institutes of Health AIDS Reagents Program) at room temperature for 1 hour followed by 2 hours of goat anti‐rabbit IgG H&L 40 nm gold (Abcam, Cambridge, England). Additional monocytes were also infected by lentiviral transduction at 1000 multiplicity of infection with 5 μg/mL polybrene (Santa Cruz Biotechnology, Santa Cruz, CA) for 24 hours and treated with HIV Tat (100 nmol/L) for 1 hour, followed by fixation in 2.5% glutaraldehyde in 0.1 mol/L sodium cacodylate buffer overnight. Both immunogold‐labeled and Tat treated monocytes were postfixated in buffered 1.0% osmium tetroxide. All coverglasses were then through a graded series of ethanol to 100% (3×) followed by graded series of 100% ethanol and hexamethyldisilazane and final series of 100% hexamethyldisilazane (3×). This last solution of 100% hexamethyldisilazane was evaporated overnight. Finally, immunogold‐labeled coverglasses were mounted onto aluminum stubs and sputter coated for 15 seconds with platinum and all other coverglasses were sputter coated with gold for 90 seconds and imaged using a Zeiss Supra 40VP Field Emission scanning electron microscope (Oberkochen, Germany).

Superresolution Confocal Microscopy Imaging

Primary monocytes were isolated as specified above, fixed using 2% paraformaldehyde, and stained for tetherin as specified for scanning electron microscopy, followed by secondary goat anti‐rabbit H&L Alexa Fluor 488 (Molecular Probes). The cellular membrane was labeled using Cell Mask Red (Invitrogen). Images were acquired using a Leica (Wetzlar, Germany) TCS SP8 confocal microscope with Leica HC PL APO CS2 ×63 magnification with 1.40 numerical aperture at room temperature. The acquisition software used was Leica LAS X (BETA) 3.5.0.18154 with Leica detectors HyD 505 to 588 nm in standard mode and HyD 654 to 731 nm in standard mode. Images were processed with Leica LIGHTNING Image Information Extraction.

Tail Bleed Assay

The time to bleeding cessation from mouse tail was performed as previously outlined. ⁵² , ⁵³ Briefly, mice over 8 weeks of age were anesthetized completely before 2 mm was cut from the tip of the tail. Time was recorded from the time of the cut to when bleeding ceased.

Statistical Analysis

All statistical analyses were performed using Prism (GraphPad Software, La Jolla, CA). Specific tests used to compare data are stated in figure legends. Statistical significance indicated as *P<0.05, **P<0.01, and ***P<0.001 shown in figures; all error bars shown in figures represent the SEM. In results, means are provided with ±SEM.

Results

Decreased Tetherin Expression on CD14+ Monocytes Correlates to Increased Levels of Circulating MDMVs During HIV Infection

Because tetherin has been demonstrated to sequester virus particles, which are of similar size and biogenesis as MVs, we investigated whether tetherin also regulates MV release. Thus, tetherin expression on the surface of monocytes was measured in 20 HIV‐positive and 20 HIV‐negative individuals (Table 1). Whole blood from donors was fixed and stained with the monocyte markers CD14 and CD16 to separate monocytes into their subsets, classical (CD14^high,CD16^low), intermediate (CD14^high,CD16^low), and nonclassical (CD14^low,CD16^high). Tetherin expression on the surface of cells was measured by flow cytometry. By both median fluorescence intensity (MFI) and percentage, tetherin expression was significantly decreased the CD14+ classical (MFI HIV−, 6091±328.3 versus MFI HIV+ mean, 4677±283.4; P=0.0009) (percent HIV−, 84.72±4.226 versus percent HIV+, 70.58±4.483; P=0.0471) and intermediate CD14+ subsets (MFI HIV−, 5947±256.2 versus MFI HIV+, 4829±262.1; P=0.0114) (percent HIV−, 82.9±3.135 versus percent HIV+, 55.21±5.479; P<0.0001), but not in the CD14− nonclassical subset (MFI HIV−, 2800±280.4 versus MFI HIV+, 2326±170.5; P=0.5130) (percent HIV−, 27.74±4.166 versus percent HIV+, 13.94±2.247; P=0.0547) (Figure 1A and 1B). Such a decrease was continually observed in combined population of CD14+ monocytes derived from HIV‐infected individuals (MFI HIV−, 5961±332.5 versus MFI HIV+, 4664±260.3; P=0.0039) (percent HIV−, 89.02±3.597 versus percent HIV+, 72.45±4.187; P=0.0007) (Figure 1C and 1D). Patient characteristics, including, sex (male, 5209±283 versus female, 5484±412.9; P=0.5742), age (MFI, 3.709±0.01946 versus age, 1.685±0.01615; P=0.349), race/ethnicity (White, 4962±261.7 versus Hispanic, 5848±930.5; P=0.6718; or versus Black, 5209±468.1; P=0.8771; Hispanic versus Black; P=0.8277), viral load (undetectable, 4693±298.1 versus detectable, 4496±494.4; P=0.7947), CD4+ T‐cell count (MFI, 3.657±0.0233 versus CD4, 2.919±0.02716; P=0.071, and specific cART regimen (nucleoside reverse transcriptase inhibitor [NRTI]/protease inhibitor [PI], 4850±502.2 versus NRTI/nonnucleoside reverse transcriptase inhibitor [NNRTI], 4187±277.5; P=0.5407; or versus NRTI/integrase strand transfer inhibitors [INSTI], 4571±472.7; P=0.8935; NRTI/NNRTI versus NRTI/ISNTI; P=0.8212) did not influence tetherin levels (Figure S1A).

Figure 1. Decreased surface tetherin levels on CD14+ monocytes correlate with increased circulating CD14+ microvesicles during HIV infection.

A through D, Flow cytometric analysis of surface tetherin expression on monocytes in HIV− (N=20) and HIV+ (N=20) donors. Monocytes were separated into their subtypes by the markers CD14 and CD16, and levels of tissue factor (TF) were quantified. Tetherin levels are reported by median fluorescence intensity (MFI; A and C) or percent (B and D). Two‐way ANOVA test with Sidak's multiple comparison test was used to compare subsets between HIV− and HIV+ (A and B). The CD14+ classical and intermediate subsets are shown pooled together (C and D). An unpaired t test was used to compare MFI (C) and the Mann–Whitney test was used to compare percentage (D) of tetherin between HIV− and HIV+. E through G, Circulating microvesicles in donor plasma was quantified by volumetric flow cytometry quantification in HIV− (N=20) and HIV+ (N=18) donors. Microvesicles expressing tissue factor (TF, E), CD14 (F), or both TF and CD14 (G) were enumerated. Unpaired t tests were performed to compare microvesicles between HIV− and HIV+. H through J, Pearson analysis was performed to evaluate the correlation between the amount of CD14+ MVs to tetherin levels, in terms of both MFI and percent (N=38), *p<0.05, **p<0.01, ***p<0.001 and ****p<0.0001.

Circulating microvesicles were quantified from the same individuals as above. Staining for CD14 and TF allowed enumeration of inflammatory MDMVs using flow cytometry. To ensure the authenticity of CD14+ only microvesicles, the blood samples were stained with 2 separate CD14‐specific antibodies that were raised against distinct epitopes and were distinguished by different fluorophores. Our analyses showed that the estimation of microvesicles in a given sample was comparable regardless of antibodies used (fluorescein isothiocyanate, 1±0.2739 versus phycoerythrin, 1.022±0.1954; P<0.0001) (Figure S2A). Furthermore, the number of CD14+ microvesicles from each donor correlated when 2 different instruments were used to enumerate (Nanosight, 1.003±0.03917 versus Accuri, 0.9785±0.1975; P=0.0500) (Figure S2B).

When microvesicles carrying multiple markers were quantified, HIV‐infected individuals were found to contain increased levels of circulating TF+ (HIV−, 4.871±0.07979 versus HIV+, 5.121±0.08974; P=0.0441) or CD14+ microvesicles (HIV−, 5.101±0.07681 versus HIV+, 5.412±SEM 0.08396; P=0.0095) Figure 1E and 1F), in agreement with previous studies. ³³ , ⁵⁴ Furthermore, CD14+ TF+ microvesicles were also increased in HIV‐infected individuals (HIV−, 4.466±0.06587 versus HIV+, 4.79±0.09423; P=0.0068) (Figure 1G). In this cohort, sex (TF+ male, 5.451±0.08348 versus TF+ female, 5.338±0.1033; P=0.4022) (CD14+ male, 5.278±0.07124 versus female, 5.203±0.1133; P=0.3249) (CD14+ TF+ male, 4.637±0.07684 versus CD14+ TF+ female, 4.593±0.106; P=0.4919), age (age, 1.684±0.01697 versus TF+ 5.406±0.06469; P=0.589; or versus CD14, 5.248±0.06149; P=0.206; or versus CD14+ TF+, 4.62±0.06177; P=0.220), race/ethnicity (TF+ White, 5.408±0.07736 versus TF+ Hispanic, 5.279±0.1133; P=0.9032; or versus TF+ Black, 5.43±0.15; P=0.9892; TF+ Hispanic versus TF+ Black; P=0.8846) (CD14+ White, 5.258±0.08142 versus CD14+ Hispanic, 5.041±0.07638; P=0.7264; or versus CD14+ Black, 5.263±0.106; P=0.9995; CD14+ Hispanic versus CD14+ Black; P=0.7378) (CD14+ TF+ White, 4.63±0.07693 versus CD14+ TF+ Hispanic, 4.306±0.04846; P=0.4880; or versus CD14+ TF+ Black, 4.657±0.1175; P=0.9816; CD14+ TF+ Hispanic versus CD14+ TF+ Black; P=0.4773) CD4 cell (CD4 2.926±0.02949 versus TF+ 5.543±0.08299; P=0.077; or versus CD14+ 5.412±0.08396; P=0.375; or versus CD14+ TF+, 4.79±0.09423; P=0.929) or viral counts (TF+ undetectable, 5.551±0.0977 versus TF+ detectable, 5.503±0.127; P=0.8346) (CD14+ undetectable, 5.39±0.09738 versus CD14+ detectable, 5.523±0.142; P=0.4265) (CD14+ TF+ undetectable, 4.752±0.1106 versus CD14+ TF+ detectable, 4.984±0.05132; P=0.3754) did not significantly influence CD14+ microvesicle (Figure S1B), TF+ microvesicles (Figure S1C), or CD14+ TF+ microvesicle levels (Figure S1D). CD14+ TF+ did not differ in regards to cART regime (NRTI/PI, 4.887±0.09555 versus NRTI/NNRTI, 4.555±0.1166; P=0.3229; or versus NRTI/integrase strand transfer inhibitor [INSTI], 4.979±0.2113; P=0.9016; NRTI/NNRTI versus NRTI/INSTI; P=0.1724); however, CD14+ (NRTI/PI, 5.428±0.1114 versus NRTI/NNRTI, 5.156±0.1334; P=0.3520; or versus NRTI/INSTI, 5.658±0.1472; P=0.4345; NRTI/NNRTI versus NRTI/INSTI; P=0.0477) and TF+ MVs (NRTI/PI, 5.644±0.1643 versus NRTI/NNRTI, 5.222±0.1249; P=0.0890; or versus NRTI/INSTI, 5.736±0.07388; P=0.8593; NRTI/NNRTI versus NRTI/INSTI; P=0.0354) did significantly differ when different types of cART were compared (Figure S1B and S1C). However, this is unlikely because of direct effects from integrase inhibitors themselves, as primary monocytes treated with the integrase inhibitor raltegravir [RAL] did not increase expression of tissue factor (not treated [NT], 6.016±1.57 versus dimethyl sulfoxide, 7.28±0.1168; P=0.8058; or versus RAL, 4.202±1.06; P=0.3527; or versus RAL+ Tenofovir disoproxil fumerate [TDF] + Emtricitabine [FTC], 4.87±1.141; P=0.2165) or release of microvesicles (NT versus dimethyl sulfoxide, 0.8878±0.1355; P=0.9517; or versus RAL, 0.9489±0.1673; P=0.9947; or versus RAL+ TDF+FTC, 0.9291±0.3123; P=0.9865) (Figure S3).

Next, in an effort to investigate if there was a link between microvesicle quantities and tetherin levels on CD14+ monocytes, statistical correlation analyses were performed. TF+ microvesicles did not correlate with levels of tetherin on CD14+ monocytes (TF+ microvesicles, 5.406±0.06469 versus MFI, 5367±241.3; P=0.672; or versus percent, 81.52±3.12; P=0.584) (Figure 1H), likely because other cells besides monocytes can contribute to this population of microvesicles. ²⁴ , ⁵⁵ However, CD14+ tetherin levels were found to be negatively correlated with both CD14+ (5.248±0.06149 versus MFI, 5367±241.3; P=0.065; or versus percent, 81.52±3.12; P=0.032) and CD14+ TF+ microvesicle levels (4.62±0.06177 MFI, 5367±241.3; P=0.021; or versus percent, 81.52±3.12; P=0.005) (Figure 1I and Figure 1J). These data demonstrate that there is a direct relationship between tetherin expression on the surface of monocytes and MDMV release in HIV‐infected subjects.

Tetherin is Colocalized With Microvesicles on the Surface of Monocytes

Because of the correlation between tetherin levels and MDMV production in monocytes, we next determined whether tetherin was physically localized next to microvesicles on monocytes. Primary monocytes were isolated from healthy HIV‐negative donors and treated with 20 ng/mL tumor necrosis factor alpha to induce microvesicle formation. Cells were fixed, immunogold labeled for tetherin, and processed for electron microscopy. Consistent with scanning electron microscopy imaging of extracellular vesicles reported in other cells, ⁵⁶ , ⁵⁷ vesicles of similar size and characteristics as microvesicles were observed on the surface of monocytes. Tetherin was found to be localized close to these microvesicles (Figure 2A, arrows). These results were further verified by performing superresolution confocal microscopy (Figure 2B; Figure S4). The images acquired by these methods also show that tetherin is localized directly at the base of microvesicle‐sized vesicles on the surface of monocytes, overall demonstrating that tetherin is in the correct position to keep microvesicles arrested on the cell surface.

Figure 2. Tetherin localizes at the site of budding microvesicles on the surface of monocytes.

A, Primary monocytes were isolated from HIV− donors followed by treated with 20 ng/mL tumor necrosis factor alpha overnight to induce microvesicle secretion. Cell were then fixed and stained using immunogold labeling for tetherin, located next to budding microvesicles (arrows). Red bar in left image denotes 1 μmol/L, and red bar in right image denotes 200 nm. B, Primary monocytes were isolated as above, fixed, and stained for tetherin and the cell membrane was labeled with Cell Mask Red. Tetherin again appeared next to budding microvesicles on the cell membrane (arrows). White bars on top images denote 5 μm and white bar on bottom images denote distances as stated.

Exposure to HIV‐Encoded Tat Decreases Tetherin Surface Expression on Primary Monocytes With Concomitant Increase in Microvesicle Production

Tetherin is well known to be downregulated by the HIV viral protein Vpu. ³⁴ However, Vpu is a late‐expressed protein and is not secreted outside the cell. ⁵⁸ Hence, it is extremely unlikely that monocytes, which rarely harbors HIV in subjects on suppressive cART, ⁵⁹ , ⁶⁰ will be exposed to Vpu in aviremic individuals. This led us to investigate if other viral proteins could affect tetherin expression. The HIV viral protein Tat is an early expressed protein that is secreted by infected cells in detectable amounts in the serum of cART‐treated individuals. ⁶¹ , ⁶² Furthermore, cells (including monocytes) exposed to Tat have been shown to exhibit a wide variety of changes at the transcriptional level. ⁶³ To investigate whether Tat could be modulating tetherin expression, we isolated primary monocytes from HIV‐negative donors and exposed them to Tat. Short 4‐hour exposures to Tat did not affect tetherin expression (NT, 50.63±16.58 versus Tat, 57.43±17.93; P=0.2328) (Figure S5A); however, Tat treatment of monocytes for 18 hours significantly decreased cell surface tetherin expression (MFI NT, 9624±424 versus Tat, 4862±634; P=0.0006) (percent NT mean, 88.1±3.133 versus Tat, 46.67±9.204; P=0.0011) (Figure 3A) (NT, 54.38±7.851 versus 500 nmol/L 29.01±4.259; P=0.0240, or versus 100 nmol/L 35.93±3.914; P=0.1119; or versus 50 nmol/L 37.83±5.581; P=0.1670; or versus 5 nmol/L 49±9.256; P=0.8968) (Figure S5B) as well as at the mRNA level (0.19±0.07506; P=0.0004) (Figure 3B). In addition to microvesicle quantification, we also sought to quantify expression of TF on monocytes in response to Tat because of TF+ microvesicles' ability to augment coagulation. ²⁴ , ⁶⁴ , ⁶⁵ Furthermore, monocytes during HIV infection, even while patients are undergoing cART, express high levels of TF, ⁶⁶ and thus, we investigated whether Tat could be contributing to this observed increase in expression. Surface TF expression increased on monocytes exposed to Tat (MFI NT, 694.1±12.12 versus Tat, 963.4±26.99; P=0.0006) (percent NT, 28.89±2.95 versus Tat, 47.21±2.355; P=0.0003) (Figure 3C).

Figure 3. HIV Tat decreases levels of tetherin on the surface of monocytes and increases tissue factor and microvesicle release.

Primary monocytes from HIV− donors were exposed to the HIV viral protein Tat (500 nmol/L) for 18 hours. A and C, Cells were then fixed and levels of tetherin (A) or tissue factor (C) on the surface of the cells were measured by flow cytometry and compared to cells that were not treated (NT). Expression levels were compared using either MFI followed by analysis with a Mann–Whitney test or percent expression followed by unpaired t test comparison. B, mRNA was isolated from NT or Tat‐treated monocytes and the levels of tetherin mRNA transcripts were compared to the housekeeping gene GAPDH. Unpaired t test was used to compare NT to Tat expression levels of mRNA. D, The number of microvesicles released from monocytes was quantified by the NS300 and Nanosight Tracking Analysis with 5 videos each lasting 30 seconds. Unpaired t test was used to compare NT to Tat. Data for all experiments in this figure were performed with 3 biological replicates with 2 to 3 technical replicates per group. In (B) technical replicates were averaged per biological sample. In (A, C, and D) all data points for each technical replicate from 3 biological replicates were included (N=7–9), **p<0.01, ***p<0.001.

To investigate whether Tat also increased MDMV release, media were collected from Tat treated monocytes, and the number of released vesicles was enumerated. Microvesicles were quantified by NanoSight Tracking Analysis, measuring 200 to 1000 nm in diameter to exclude exosomes. ⁶⁷ Upon Tat treatment, monocytes significantly increased release of MDMVs into the media (nontreated, 1±0.07099 versus Tat, 1.758±0.205; P=0.003) (Figure 3D). MDMVs from nontreated and Tat‐treated monocytes were of similar size in both mean (nontreated, 279.8±9.157 versus Tat, 283.4±5.34; P=0.6477) and mode (nontreated, 238±0.8602 versus Tat, 237.3±2.106; P=0.7431) diameter (Figure S6). Additionally, to investigate whether microvesicles were being quantified on the NS300 or if it was analyzing erroneous matter in the media, media from monocytes incubated overnight was compared with the RPMI in which the monocytes were suspended. Media from cells contained significantly more microvesicles as measured by calcein green, demonstrating these microvesicles were membrane enclosed (RPMI, 14 400±3150 versus Cell Media, 71 511±5389; P=0.0008) (Figure S7A). Additionally, there were more microvesicles in the media from cells that had the microvesicle marker Annexin V alone (RPMI, 18 371±2605 versus Cell Media, 137 784±31 781, P=0.0200) (Figure S7B) or in combination with the monocyte marker CD14 (RPMI, 1788±384 versus Cell Media, 6892±903.6; P=0.0065) (Figure S7C). Taken together, monocytes treated with HIV Tat in vitro corroborated with monocytes isolated from HIV‐positive subjects in respect to their (1) decreased surface tetherin expression, (2) increased release of MDMVs, and (3) increased surface TF expression.

The relationship between tetherin and MDMVs was further examined by performing electron microscopy. Monocytes were first infected with a lentivirus vectors either containing a cassette expressing wild‐type tetherin or nothing (empty vector). After infection, cells were treated with HIV Tat to observe morphological changes.

Under normal conditions, monocytes display large ruffles (lamellipodia) and microvesicles on their surface (Figure 4A and 4B, vector untreated, arrowheads; Figure S8). Monocytes overexpressing tetherin wild‐type (WT) exhibited a morphology wherein the cells displayed abundant tubular structures (pseudopodia, Figure 4B, WT untreated, asterisks). Further, protrusions that match microvesicles in both size and morphology were located on the ends of pseudopodia (Figure 4B, WT, arrowheads, asterisk), suggesting these structures are where biogenesis and release of microvesicles occurs. These data align with other reports where microvesicles have been found on the ends of pseudopodia in scanning electron microscopy images of platelet‐derived microvesicles. ⁶⁸ , ⁶⁹ These microvesicle structures disappeared upon treatment with Tat (Figure 4B, vector and WT Tat‐treated). These images visualize the close relationship of tetherin overexpression with extended pseudopodia capped with microvesicles. These data support the direct connection in Tat‐mediated microvesicle release through tetherin downregulation demonstrated in Figure 3.

Figure 4. Scanning electron microscopy morphology of monocytes overexpressing tetherin.

Primary monocytes were isolated and infected with lentiviral vector to over express tetherin. Either an empty vector (Vector) or wild‐type (WT) tetherin. Cells were either left untreated or exposed to a low level of HIV Tat (100 nmol/L) for 1‐hour prior fixation and preparation for scanning electron microscopy imaging. (A) Monocytes displayed typical morphology, with large ruffles and protrusions. Red bar denotes 2 μm. (B) Microvesicle‐sized and shaped vesicles (red arrows) disappeared from the surface of Tat‐treated cells. Tubular structures below vesicles (red asterisk) are visualized upon tetherin overexpression. Red bars denote 200 nm.

Tetherin Deficient Mice Exhibit Higher Numbers of Circulating MDMVs and Reduced Tail Bleeding Time

To further investigate the direct role of tetherin in microvesicles sequestration, we collected blood from a retro‐orbital puncture of tetherin‐deficient knockout mice, as well as age‐ and sex‐matched C57/BL6 mice that expressed wild‐type tetherin. Tetherin knockout mice had significantly higher levels of circulating MDMVs in blood compared with their WT counterparts (WT, 520 873±182 698 versus KO, 1 007 040±101 256; P=0.0356; Figure 5A). These data demonstrate the direct role of tetherin in MDMV sequestration.

Figure 5. Mice lacking tetherin increase monocyte‐derived microvesicles (MDMV) release and take less time for bleeding cessation.

A, Blood was taken from age and sex matched wild‐type (WT) or tetherin knockout mice via retro‐orbital bleed (N=10–11 per group). Plasma was isolated by centrifugation, stained for CD115+, and MDMVs were enumerated by flow cytometry. Unpaired t test was performed to compare number of MDMVs between WT and knockout mice. B, Tail bleed measurements were taken from WT or tetherin knockout mice (N=6 per group) and compared using unpaired t test. The time for bleeding cessation was measured from the beginning of the incision to when bleeding ceased, *p<0.05.

MDMVs have been shown to play a significant role in coagulation, and thus we measured the time required for the bleeding cessation in mice via tail bleed assays. Bleeding cessation occurred rapidly in tetherin knockout mice, as compared with tetherin wild‐type mice, with concomitant changes in the number of circulating MDMVs (WT, 200.5±39.06 versus KO, 97.33±13.4; P=0.0316) (Figure 5B). It is unlikely that platelet activation is influencing the faster bleeding cessation for tetherin knockout mice since their platelets contain the same levels of activation compared with their wild‐type counterparts (WT, 4.467±0.3844 versus KO, 3.35±0.75; P=0.2321; Figure S9). The data presented here strongly suggest that tetherin directly affects MDMV release, which can lead to an increased procoagulatory state.

Discussion

Despite the success of cART in controlling viral loads, individuals infected with HIV still suffer from numerous inflammatory diseases that are greatly augmented by monocyte activation. ¹ , ² , ³ , ⁴ , ¹⁰ Monocytes exacerbate inflammation by releasing microvesicles; however, the mechanism by which microvesicles are released has remained largely unknown. In this current study, we identified a novel mechanism of microvesicle regulation through the surface protein tetherin, which plays an integral role in restricting microvesicle budding from monocytes during HIV infection. Our results demonstrated a direct correlation between decreased tetherin levels and increased circulating MDMVs in both HIV clinical samples and a tetherin knockout mouse model. Furthermore, we demonstrated that the HIV viral protein Tat decreases tetherin levels and increases microvesicle release from monocytes, suggesting that Tat contributes to the decreased tetherin levels observed in HIV individuals.

Our study further established that HIV Tat increases in vitro TF expression on monocytes, while our clinical data showed an increase of TF+ MDMVs in the circulation of HIV‐infected individuals. The increase in TF+ microvesicle levels is particularly important during inflammation because of TF's promotion of the extrinsic coagulation cascade leading to clot‐forming fibrin and platelet activation. ⁷⁰ Indeed, previous studies have stressed the importance of TF+ microvesicles by demonstrating that TF expression bolsters microvesicles' ability to induce coagulation and increase thrombus size. ²⁴ , ⁶⁴ , ⁶⁵ Concurrent with these findings, we demonstrated that tetherin knockout mice had increased circulating levels of MDMVs that functionally resulted in rapid bleeding cessation. These procoagulatory TF+ MDMVs have important implications during HIV infection, as HIV‐positive individuals have higher incidences of coronary heart disease ⁷¹ and myocardial infarctions. ⁷² Understanding the impact and regulation of TF+ MDMVs could prove pivotal for developing new treatments.

The discovery that Tat can regulate tetherin is pertinent since circulating Tat remains present in the blood of HIV‐infected individuals receiving cART treatment ⁶¹ , ⁶² and thus can mediate inflammation. Previous studies have also thoroughly established Tat as a proinflammatory protein that regulates cellular protein expression through a variety of mechanisms. For example, Tat translocates through the nuclear membrane ⁷³ and binds to nuclear factor kappa light‐chain enhancer of activated B cells promoter elements to downregulate the expression of host proteins. ⁷⁴ The tetherin promoter contains nuclear factor kappa light‐chain enhancer of activated B cells–responsive elements, ⁷⁵ and thus could be occupied by Tat, thereby decreasing tetherin transcription. Alternatively, Tat has been shown to activate the MyD88 immune signaling pathway, ⁷⁶ which negatively regulates tetherin. ⁷⁷ Ultimately, identifying the mechanism by which Tat regulates tetherin will provide insight into how HIV evades host antiviral response and may indicate another potential target for future therapies.

Our studies indicate a novel mechanism of inhibiting MDMV release by tetherin, a protein that act as a physical tether and is already highly documented to restrict virus particle release. ⁷⁸ , ⁷⁹ , ⁸⁰ , ⁸¹ Microvesicles and virus particles have similar biogenesis, both emerging from the plasma membrane at lipid rafts where tetherin's glycosylphosphatidylinositol anchor is localized, ⁸² suggesting that tetherin could act as a physical tether for MDMVs as well. Our data indeed demonstrate that tetherin was appropriately localized at the cell membrane to physically sequester MDMVs. Further, cells overexpressing tetherin had increased pseudopodia with microparticle‐sized vesicles on the ends. Monocytes could feasibly express smaller versions of these tetherin‐induced pseudopodia located directly below microvesicles, consequently tethering these microvesicles to the surface. Despite the strong plausibility of physically restricting microvesicles, tetherin is a multifunctional protein and could also regulate MDMVs by other mechanisms crucial to microvesicle release, such as by cytoskeleton and protein scaffolding organization, ³⁷ , ⁸³ , ⁸⁴ lipid raft arrangement, ⁸⁵ , ⁸⁶ or nuclear factor kappa light‐chain enhancer of activated B cells activation. ⁸⁷ , ⁸⁸

Furthermore, it is also important to note microvesicle formation is influenced by multiple mechanisms, including lipid rearrangement, ⁸⁹ actin and cytoskeleton organization, ⁸³ , ⁹⁰ cytosolic calcium concentrations, ⁹¹ and recruitment of cellular proteins. ⁹² Because of the complex interplay of these multiple factors contributing to microvesicle regulation, the exact mechanisms tetherin uses to regulate microvesicles, and its overall impact could be context dependent. Further research is necessary to fully elucidate the mechanism by which tetherin regulates MDMVs and its specific impact in different diseases and situations.

Although MDMVs' contribution to inflammation has been well established throughout the literature, little is known about their regulation or role during HIV infection. Our findings describe a novel mechanism of MDMV restriction by the membrane protein tetherin. In HIV clinical samples, decreased tetherin expression on monocytes correlated with increased circulating MDMVs, and a murine knockout tetherin demonstrated increased circulating MDMVs and increased bleeding cessation. Furthermore, our research revealed that exposure to the viral protein Tat could be a probable mechanism for the decreased tetherin and increased circulating TF+ MDMVs observed in HIV‐infected individuals. Collectively, this work builds a strong foundation to further investigate a novel mechanism for MDMV regulation, which might be expanded to all microvesicle populations. Thus, targeting tetherin could lead to new treatments for many inflammatory diseases related to microvesicles, both in the HIV‐infected and general populations.

Sources of Funding

The authors acknowledge funding for this work from the U.S. National Institutes of Health (R01 HL128155, R01 HL123346, R01 NS066801 and T32 AI049815) and also the University of Rochester Center for AIDS Research (UR‐CFAR; P30 AI078498) for their support of core facilities.

Disclosures

None.

Supporting information

Figures S1–S9

(J Am Heart Assoc. 2020;9:e015998 DOI: 10.1161/JAHA.120.015998.)