NETosis Drives Blood Pressure Elevation and Vascular Dysfunction in Hypertension

Abstract

BACKGROUND:

Neutrophil extracellular traps (NETs) are composed of DNA, enzymes, and citrullinated histones that are expelled by neutrophils in the process of NETosis. NETs accumulate in the aorta and kidneys in hypertension. PAD4 (protein-arginine deiminase-4) is a calcium-dependent enzyme that is essential for NETosis. TRPV4 (transient receptor potential cation channel subfamily V member 4) is a mechanosensitive calcium channel expressed in neutrophils. Thus, we hypothesize that NETosis contributes to hypertension via NET-mediated endothelial cell (EC) dysfunction.

METHODS:

NETosis-deficient Padi4^−/− mice were treated with Ang II (angiotensin II). Blood pressure was measured by radiotelemetry, and vascular reactivity was measured with wire myography. Neutrophils were cultured with or without ECs and exposed to normotensive or hypertensive uniaxial stretch. NETosis was measured by flow cytometry. ECs were treated with citrullinated histone H3, and gene expression was measured by quantitative reverse transcription PCR. Aortic rings were incubated with citrullinated histone H3, and wire myography was performed to evaluate EC function. Neutrophils were treated with the TRPV4 agonist GSK1016790A. Calcium influx was measured using Fluo-4 dye, and NETosis was measured by immunofluorescence.

RESULTS:

Padi4^−/− mice exhibited attenuated hypertension, reduced aortic inflammation, and improved EC-dependent vascular relaxation in response to Ang II. Coculture of neutrophils with ECs and exposure to hypertensive uniaxial stretch increased NETosis and accumulation of neutrophil citrullinated histone H3. Histone H3 and citrullinated histone H3 exposure attenuates EC-dependent vascular relaxation. Treatment of neutrophils with the TRPV4 agonist GSK1016790A increases intracellular calcium and NETosis.

CONCLUSIONS:

These observations identify a role of NETosis in the pathogenesis of hypertension. Moreover, they define an important role of EC stretch and TRPV4 as initiators of NETosis. Finally, they define a role of citrullinated histones as drivers of EC dysfunction in hypertension.

Novelty and Significance.

What Is Known?

• Hypertension is associated with inflammation, and neutrophils are the most abundant leukocyte in circulation in humans.

• Neutrophils can undergo NETosis, a process whereby the neutrophil expels DNA and nuclear components and neutrophil extracellular traps (NETs) accumulate in peripheral vascular tissue in hypertension.

• TRPV4 is a mechanosensitive calcium channel that is expressed in neutrophils.

What New Information Does This Article Contribute?

• NETosis contributes to blood pressure elevation and endothelial cell dysfunction in hypertension.

• H3-Cit, a product of NETosis, activates endothelial cells and disrupts endothelial-dependent vascular function.

• The mechanosensitive calcium channel TRPV4 (transient receptor potential cation channel subfamily V member 4) induces NETosis in isolated neutrophils.

Hypertension is the most important driver of cardiovascular disease and is driven by inflammation. A better understanding of the pathophysiology of inflammation in hypertension will support the development of durable therapies. Neutrophils are the most abundant circulating leukocyte in humans and can undergo NETosis, a process whereby nuclear contents, including DNA and histones, are expelled. This process requires the calcium-dependent enzyme PAD4. The current study describes a role of NETs as drivers of hypertension and vascular dysfunction. NETosis-deficient mice exhibit attenuated hypertension and improved endothelial cell-dependent vascular function in response to Ang II (angiotensin II). Hypertensive stretch of endothelial cells increases NETosis, and endothelial cells support neutrophil survival. Hypertensive stretch leads to the accumulation of citrullinated histone H3 (H3-Cit) in neutrophils. Treatment of endothelial cells with H3-Cit leads to endothelial cell activation. Exposure of aortic rings to histone H3 or H3-Cit disrupts endothelial cell function. Finally, agonism of the mechanosensitive calcium channel TRPV4 in neutrophils increases intracellular calcium and induces NETosis. These findings describe a critical role of NETosis as a driver of hypertension. Moreover, they suggest that targeting NETosis may be an important therapeutic approach for the treatment of hypertension and resultant cardiovascular diseases.

Meet the First Author, see p 1401

Hypertension is the most important modifiable risk factor for cardiovascular disease.¹ Treatment of hypertension is the most common reason for the use of prescription medications and office visits in the United States.^2,3 The prevalence of hypertension in the United States was 45.4% in 2017 to 2018 according to NHANES (National Health and Nutrition Examination Survery) data, and the control of blood pressure is poor with control rates <50%.^4–6 Moreover, blood pressure control is declining particularly among adults over 75 years old, women, and non-Hispanic Black adults.⁶ There remains a significant need for new and durable therapies for the treatment of hypertension.

Hypertension is associated with immune activation.⁷ Both cells of the innate and adaptive immune systems play a role in immune activation and inflammation in hypertension.⁸ Neutrophils play an important role in the innate immune response and are the most abundant leukocyte in circulation in humans.⁹ Neutrophils exert their function by engulfing pathogens and releasing enzymes that contribute to the destruction of bacteria and viruses.⁹ In human hypertension, neutrophil count is increased.¹⁰ The neutrophil to lymphocyte ratio is also increased in patients with hypertension, and this increase correlates with the risk of developing hypertension.^10,11 In murine hypertension, neutrophils are released from depot tissues and migrate into peripheral tissues.¹² Neutrophils also respond to pathogens by forming neutrophil extracellular traps (NETs) in the process of NETosis. NETs are composed of decondensed chromatin, granular proteins, and enzymes that are antimicrobial.¹³ NETs are proinflammatory and play a role in immune activation and inflammation in many disease conditions.^13–15 NETosis in peripheral tissues such as the kidney and aorta is augmented in hypertension in both humans and rodent models.^12,16 NETs are also found in mesenteric vessels in hypertensive rat and murine models.¹⁷ Currently, however, the role of NETosis in the pathogenesis of hypertension is unclear. Moreover, the stimuli that contribute to NETosis in hypertension are poorly characterized.

NETs promote atherosclerosis and thrombosis and contribute to the activation of platelets.^18–20 However, NETs also promote these conditions in part through effects on endothelial cell (EC) activation. Exposure of NETs to ECs results in an increase in VCAM-1 and ICAM-1 expression.²¹ The specific components of NETs driving this process, however, remain unknown. Citrullinated histone H3 (H3-Cit) is a product of PAD4 (protein-arginine deiminase type-4)–mediated conversion of arginine residues to citrulline. H3-Cit has been shown to disrupt the microvascular EC barrier.²² Moreover, neutralizing H3-Cit in a mouse model of acute lung injury attenuates alveolar edema and hemorrhage.²³

NETosis is initiated by a variety of stimuli including bacterial cell wall components such as lipopolysaccharide, exposure to cytokines such as TNF (tumor necrosis factor), and immune complexes.¹³ The mechanisms driving NETosis remain unclear; however, numerous components seem to be required for this process including reactive oxygen species (ROS) generation by the NADPH oxidase, citrullination of histones by PAD4, and increased intracellular calcium.²⁴ Calcium ionophores are potent inducers of NETosis and calcium chelators such as EGTA prevent NETosis.^12,25 Calcium is required for the activation of PAD4 by inducing conformational changes that generate an active site.²⁶ The mechanosensitive calcium channel TRPV4 (transient receptor potential cation channel subfamily V member 4) is highly expressed in neutrophils. Moreover, inhibition of TRPV4 attenuates neutrophil ROS production, adhesion, and chemotaxis in a model of acute lung injury.^27,28 Treatment of neutrophils with the TRPV4 agonist GSK1016790A increases ROS production in a dose-dependent manner.²⁷ Moreover, in a model of myocardial ischemia/reperfusion injury, TRPV4 mediates ROS production and MPO (myeloperoxidase) release from neutrophils.²⁷

Based on these findings, we hypothesized that NETosis leads to EC activation and vascular dysfunction, thus contributing to hypertension. Moreover, we hypothesized that EC stretch is a driver of NETosis and H3-Cit, a product of PAD4, is a direct contributor to NETosis-induced EC dysfunction.

METHODS

Data Availability

The data that supports the findings of the study are available from the corresponding author upon reasonable request.

Animals

Eleven-week-old male C57BL/6J and B6.Cg-Padi4^tm1.1Kmow/J (Padi4^−/−) mice were obtained from The Jackson Laboratory and implanted with subcutaneous minipumps containing 490 ng/kg per min Ang II (angiotensin II; Sigma) for 28 days as previously described.²⁹ Blood pressure was measured using radio-telemeters, which were implanted at 12 weeks. Vascular reactivity studies were performed on the mesenteric arteries or thoracic aortic rings to observe vascular relaxation in the presence of acetylcholine and sodium nitroprusside, as previously described.³⁰ For experiments measuring vascular relaxation in response to histones, aortic rings were preincubated with either histone H3 (Cayman Chemical) or H3-Cit (Cayman Chemical) at 10 µg/mL for 2 hours before precontraction. Saline challenge was performed as previously described.^31,32 Animal procedures were approved by the Department of Veteran’s Affairs and Vanderbilt University Medical Center’s Institutional Animal Care and Use Committees before the performance of the studies. Mice were housed and cared for in accordance with the Guide for the Care and Use of Laboratory Animals, US Department of Health and Human Services. No animals were excluded from the study and randomization was not performed.

Flow Cytometry

Quantification of infiltrating immune cell populations was performed using flow cytometry. Mice were perfused with saline via left ventricular puncture following removal of the right atrium. Single-cell suspensions were prepared from aorta following digestion with collagenase A (Sigma) at 1 mg/mL, collagenase B (Sigma) at 1 mg/mL, and DNAse I (Sigma) at 0.1 mg/mL in PBS. Cell viability was assessed using the Zombie near-infrared (NIR) Fixable Viability Kit (BioLegend). Extracellular staining was performed with CD45 (BV750, clone 30-F11, BioLegend No. 103157), CD64 (PE, clone X4-5/7.1, BioLegend No. 139304), MERTK (Mer tyrosine kinase; APC, clone 2B10C42, BioLegend No. 151508), I-A/I-E (PerCP-Cy5.5, clone M5/114.15.2, BD Biosciences No. 562363), CD11c (BV650, clone N418, BioLegend No. 117339), CD3 (BV605, clone 17A2, BioLegend No. 100237), CD4 (FITC, clone GK1.5, BD Biosciences No. 557307), CD8a (BV421, clone 53-6.7, BioLegend No. 100738), and Ly6G (BV510, clone 1A8, BioLegend No. 127633) for 20 minutes at 4 °C. A known quantity of counting beads (Invitrogen) was added to each sample. Results were normalized using the bead count and represented as the number of cells per organ. To quantify NETs in mouse aortic EC coculture experiments, adherent cells were collected and evaluated as previously described.¹² Fluorescence minus one controls were used for all markers. Acquisition was performed using a Cytek Aurora flow cytometer, and analysis was performed using FlowJo software. Infiltrating immune cells were identified with the gating strategy represented in Figure S1.

Mouse Neutrophil Isolation

Murine neutrophils were obtained from the peritoneal cavity of mice following casein injection, as previously described.³³ In brief, C57BL/6J or Padi4^−/− mice aged 12 to 15 weeks were injected intraperitoneally with 1 mL of sterile casein solution (9% w/v in 1× PBS containing calcium and magnesium). A second casein injection was performed the following day, and mice were sacrificed 3 hours after injection. Peritoneal cells were harvested in 10 mL of sterile PBS supplemented with EDTA and then layered over 63% Percoll solution (Cytiva) and spun at 1000g with no brake for 20 minutes at room temperature. The lowermost neutrophil and red blood cell layer were washed and then incubated with 1× red blood cell lysis buffer (Invitrogen) and then resuspended in either culture medium (RPMI 1640 supplemented with L-glutamine, HEPES, and 2% FBS) or PBS.

Mouse Neutrophil Immunofluorescence and NET Quantitation

Neutrophils were seeded onto poly-L-lysine-coated coverslips in a 12-well plate at a density of 1 to 2×10⁶ cells per well in culture medium. Cells were treated with 1 nmol/L, 10 nmol/L, or 100 nmol/L of the selective TRPV4 agonist GSK1016790A (Cayman Chemical) or DMSO as a vehicle and then incubated for 30 minutes at 37 °C. Following incubation, neutrophils were fixed with 4% paraformaldehyde (Electron Microscopy Sciences) and blocked with 10% donkey serum in PBS overnight at 4 °C. Cells were stained with rabbit anti-H3-Cit antibody (Abcam No. 5103), washed, and incubated with anti-rabbit IgG AF488 (Invitrogen No. A21206). This was followed by washing and subsequent staining with AF647-conjugated myeloperoxidase (Abcam No. 252131). Coverslips were washed and mounted to slides using ProLong Gold Antifade Mountant with DAPI (Invitrogen). Images were collected with a Leica DM6000 microscope at 10× magnification with LasX software. Five fields of view were captured per slide, and NET quantitation was performed as area of colocalization of DAPI, H3-Cit, and MPO using Imaris Cell Imaging Software (Oxford Instruments). For the individual marker H3-Cit, captured images were analyzed in ImageJ (National Institutes of Health). Maxima count (prominence >10.00) for all images was normalized to the number of neutrophils plated.

Treatment with GSK1016790A and Intracellular Calcium Measurement

Mouse neutrophils were stained with cell-permeant Fluo-4-AM (Invitrogen) in staining buffer (1× PBS containing calcium and magnesium, HEPES, 1% albumin) at a final concentration of 1 μM for 30 minutes at 37 °C. Cells were then incubated in LIVE/DEAD Fixable Violet Dead Cell Stain Kit (Invitrogen) in PBS, followed by staining with Ly6G antibody (BV510, clone 1A8, BioLegend No. 127633) in staining buffer. Cells were then treated with GSK1016790A at 1, 10, or 100 nmol/L or DMSO as vehicle and incubated at room temperature for 10 minutes. Flow cytometry analysis was performed with the Cytek Aurora Analyzer. Calcium was quantified using the mean fluorescence intensity of the Fluo-4-AM staining. Cells were analyzed based on the gating strategy presented in Figure S2. Fluorescence minus one controls were used to determine gating strategies and staining efficacy.

Histone Treatment of ECs and RT-PCR

Mouse aortic ECs (Cell Biologics) were cultured in complete EC medium (Cell Biologics) to confluency in a 12-well plate. Cells were treated with recombinant histone H3 (Cayman Chemical) or recombinant citrullinated histone H3 (Cayman Chemical) at 10 µg/mL or with PBS as vehicle for 4 hours at 37 °C. RNA was extracted using Trizol (Invitrogen) according to manufacturer’s instructions. Reverse transcription was performed using the High Capacity cDNA Reverse Transcription kit (Applied Biosystems). Real-time quantitative PCR (RT-qPCR) was performed using TaqMan Universal PCR Master Mix (Applied Biosystems) and TaqMan assays (Applied Biosystems) for Icam1 (intracellular adhesion molecule 1), Vcam1 (vascular cell adhesion molecule 1), Sell, Sele, and Gapdh on the QuantStudio 7 Pro Real-Time PCR system. Briefly, after 2 minutes at 50 °C, initial denaturation was performed at 95 °C for 10 minutes, followed by 40 cycles of PCR consisting of denaturation at 95 °C for 15 seconds, and annealment at 60 °C for 1 minute. RNA samples were not DNase-treated. The ΔΔCt method was used to analyze results. Gene expression values were calculated based on the comparative Ct normalized to the expression values of Gapdh and expressed as fold change normalized to control.

H3-Cit Treatment of Stretched ECs and RT-PCR

Mouse aortic ECs (Cell Biologics) were cultured in complete endothelial cell medium (Cell Biologics) to confluency in UniFlex Collagen Type-I plates (Flexcell International). Cells were treated with recombinant citrullinated histone H3 (Cayman Chemical) at 10 µg/mL or with PBS as vehicle. Uniaxial-cyclical mechanical stretch was applied to these wells at 10% as previously described for 18 hours at 37 °C.³⁴ RNA was extracted using Trizol (Invitrogen) according to manufacturer’s instructions. Reverse transcription was performed as described above with the TaqMan assays (Applied Biosystems) for Il6 (interleukin 6), Tgfβ (transforming growth factor beta), Tnfα (tumor necrosis factor-alpha), and Gapdh on the QuantStudio 3 Real-Time PCR system. Gene expression values were calculated based on the comparative Ct normalized to the expression values of Gapdh and expressed as fold change normalized to control.

Mouse EC and Neutrophil Coculture and Immunoblot

ECs were grown to confluency on UniFlex Collagen Type-I plates (Flexcell International). ECs were used at passage number 3 to 6. Mouse neutrophils were plated into these plates at a density of 0.5 to 1.0×10⁶ cells per well. Uniaxial-cyclical mechanical stretch was applied to these plates at 5% and 10% stretch as previously described for 18 or 36 hours.³⁴ Cells were scraped and collected from wells and used for analysis of NETs by flow cytometry as previously described.¹² Immunoblot was performed on nonadherent neutrophils following 36 hours of coculture with ECs. Briefly, neutrophils were lysed with RIPA (Thermo Fisher) following the dissolution of 1 protease inhibitor cocktail tablet (1 tablet/10 mL RIPA buffer; Roche). Protein was quantitated with the BCA Protein Assay Kit according to the manufacturer’s instructions (Thermo Fisher). Ten micrograms of protein were separated on a 12% SDS-PAGE gel by electrophoresis. The SDS-PAGE gel was transferred to a nitrocellulose membrane. Immunoblot was performed with anti-histone H3 (citrulline R2+R8+R17) antibody (Abcam No. 5103) at 1:1000 and anti-histone H3 antibody (Cell Signaling No. 9715S) at 1:1000. Secondary incubation was performed at 1:5000 anti-Rabbit HRP (Thermo Fisher No. 31460). Pierce ECL blotting substrate was used for detection (Thermo Fisher). Immunoblots were imaged on the Invitrogen iBright CL1000 imaging system. ImageJ was used for quantitation.

Single-Cell Sequencing

We analyzed a single-cell data set we previously generated and published previously.^12,35 Briefly, spleens were harvested, and T cells were depleted using anti-CD3 microbeads (Miltenyi Biotec). Single-cell sequencing was performed using the Chromium Single-Cell v2 5′ Chemistry Library, Gel Bead, Multiplex, and Chip Kits (10× Genomics) according to the manufacturer’s protocol. A total of 12 000 cells were targeted per well. Libraries were then sequenced utilizing the NovaSeq 6000 platform (Illumina). Raw base call files were demultiplexed and mapped using the cell ranger single-cell gene expression v.6.0.0 software (10× Genomics). We used Seurat v4.0.0 in R v4.0.4 for data normalization, cell filtering, dimensionality reduction, clustering, and gene expression analysis using default parameters. Cells were included with RNA counts >200 and <3500, and a mitochondrial content <15%. All data presented including Padi4 expression are from Ang II–treated C57Bl/6 male mice. The R script is available upon reasonable request. Data are available on GEO Browser (GSE200588).

Statistics

All data are expressed as mean±SEM. The normality of the distribution of data was confirmed using the D’Agostino-Pearson normality test. Comparisons of 2 variables were performed with Student t test. If 1 or both distributions was/were not normal, then a Mann-Whitney U test nonparametric test was performed. Comparison among more than 2 variables was performed with 1-way ANOVA with Tukey post hoc test. To compare 2 categorical variables, a 2-way ANOVA was used with Sidak multiple comparisons test. To compare differences in blood pressure, a 2-way ANOVA with repeated measures was used on the final week of measurement. To compare differences in vascular reactivity, a 2-way ANOVA with repeated measures was used with Tukey multiple comparison test if >2 reactivity curves were included. To determine the number of animals required for radiotelemetry, a power analysis was performed based on previously published data. Replicate numbers for in vitro experiments were based upon available reagents. Experiments were performed using methods to reduce confirmation bias. Analysis of immunofluorescence data was performed using the batch function in the Imaris software suite, which evaluates each image in a blinded fashion. All animals were sacrificed in an identical fashion using a number system without genotype identification. All procedures were performed identically for all samples. Analysis of flow cytometry was performed using identical gating strategies for each sample that is applied by FlowJo software in a blinded manner.

RESULTS

Loss of PAD4 Attenuates Ang II–Induced Hypertension

To determine the expression of Padi4 in immune cell subsets from hypertensive mice, we examined single-cell sequencing data from splenocytes harvested from mice treated with Ang II. Twenty-three clusters were identified (Figure S3A). Clusters were evaluated based upon cluster-specific gene expression to identify cell type (Figure S3B). Padi4 expression is enriched in neutrophil clusters and minimally expressed by other cell types (Figure S4A and S4B). To determine the role of PAD4 in hypertension, we treated Padi4^−/− or C57Bl/6 control mice with 490 ng/kg per minute Ang II for 4 weeks. Systolic, diastolic, and mean arterial pressure were measured by radiotelemetry. Padi4^−/− mice develop hypertension during week 1 and exhibit a reduction in systolic, diastolic, and mean arterial pressure during the following 3 weeks (Figure 1A through 1C). No difference in heart rate was observed (Figure 1D).

Figure 1.

Neutrophil extracellular trap (NET)-deficient mice exhibit attenuated hypertension. Padi4^−/− and C57Bl/6 mice were treated with Ang II (angiotensin II) infusion for 4 weeks. Radiotelemetry was used to measure (A) systolic blood pressure, (B) diastolic blood pressure, (C) mean arterial pressure (MAP), and (D) heart rate. Single day variability represents day/night cycles. Data were analyzed by repeated-measures 2-way ANOVA (5 animals per group). BP indicates blood pressure.

Padi4^−/− Mice Exhibit Improved EC-Dependent Vascular Relaxation and a Reduction in Aortic Inflammation

To evaluate the role of NETosis on vascular reactivity, mesenteric arteries were harvested and vascular function was determined by wire myography. No difference in response to sodium nitroprusside was observed; however, Padi4^−/− mice exhibit improved vascular relaxation in response to acetylcholine (Figure 2A and 2B). To evaluate renal sodium and volume retention, mice received an injection of normal saline equal to 10% of their body weight following Ang II infusion, and urine volume was measured over 4 hours as previously described.^31,32 Padi4^−/− mice exhibit a trend for improvement in renal sodium and volume excretion (Figure 2C). Aortas were harvested following Ang II infusion, and inflammation was evaluated by flow cytometry using the gating strategy presented in Figure S1. Neutrophil, dendritic cell, and cytotoxic CD8⁺ T cell infiltration is markedly reduced in Padi4^−/− mice compared with C57Bl/6 mice (Figure 2D through 2F).

Figure 2.

Neutrophil extracellular trap (NET)-deficient mice exhibit improved endothelial cell function and reduced vascular inflammation in response to Ang II (angiotensin II). Padi4^−/− and C57Bl/6J mice were treated with Ang II for 4 weeks. Wire myography was used to measure vascular relaxation of mesenteric vessels in the presence of (A) sodium nitroprusside (SNP) and (B) acetylcholine (ACH). Data were analyzed by repeated-measures 2-way ANOVA (4 animals per group). C, Urine volume was quantitated following a saline challenge (4 C57Bl/6 animals and 5 Padi4^−/− animals). Aortas were harvested and quantitation of (D) neutrophils, (E) dendritic cells (DCs), and (F) CD8⁺ cytotoxic T cells was performed by flow cytometry (10 animals per group). Data were analyzed by 2-tailed Student t test (urine volume and neutrophils/aorta) and 2-tailed Mann-Whitney U test (DCs/aorta and CD8⁺ T cells/aorta).

Coculture of Neutrophils With ECs and Exposure to Hypertensive Stretch Increase NETosis

Primary C57Bl/6 neutrophils were harvested and cultured with or without C57Bl/6 mouse aortic ECs for 18 hours. The purity of neutrophils following harvest was >90% (Figure S5). Cultured cells were exposed to normotensive (5%) or hypertensive (10%) uniaxial stretch. In the absence of ECs, 10% stretch increases NETosis. However, we found that following coculture, neutrophils exhibit increased NETosis when exposed to ECs at either 5% or 10% stretch when compared with neutrophils cultured alone. A 10% stretch augments NETosis in coculture in an additive fashion when compared with 5% stretch or neutrophils cultured alone (Figure 3A and 3B). Padi4^−/− neutrophils cocultured with ECs at 10% stretch resulted in minimal extracellular DNA excretion and NET formation compared with C57Bl/6 neutrophils (Figure S6).

Figure 3.

Hypertensive stretch augments NETosis in the presence of endothelial cells. Murine neutrophils were cultured with and without endothelial cells (ECs) under normal (5%) and hypertensive (10%) stretch for 18 hours. A, Representative flow cytometry plots of neutrophil extracellular trap (NET) formation. B, Quantitation of NETs. Data were analyzed by 2-way ANOVA with Šidák multiple comparison test (−EC+5%, n=8; −EC+10%, n=8; +EC+5%, n=6; +EC+10%, n=8). MPO indicates myeloperoxidase.

10% Stretch Increases Suicidal NETosis and Coculture of Neutrophils With ECs Increases Neutrophil Survival

Neutrophils were cocultured with ECs for 36 hours. Importantly, coculture in the presence of ECs markedly increased cell survival when compared with the culture of neutrophils alone (Figure 4A and 4B). We determined the effect of stretch in the presence of ECs at this time and found that 10% stretch increased NETosis in coculture compared with 5% stretch (Figure 5A and 5B). Evaluation of cell viability revealed an increase in suicidal NETosis at 10% stretch compared with 5% stretch (Figure 5C through 5E).

Figure 4.

Endothelial cells (ECs) prolong neutrophil survival. Murine neutrophils were cocultured with or without ECs and subjected to 5% and 10% stretch for 36 hours. Neutrophils were stained with live/dead cell stain that indicates cell viability. A, Representative histograms of live/dead in neutrophils. B, Quantitation of live neutrophils as a percent of total seeded neutrophils. Data were analyzed by 2-way ANOVA with Šidák multiple comparison test (n=12 for each group).

Figure 5.

Long-term hypertensive stretch augments suicidal NETosis in the presence of endothelial cells. Murine neutrophils were cocultured with endothelial cells and subjected to normotensive (n=12) and hypertensive (n=12) stretch for 36 hours. A, Representative flow cytometry plots of neutrophil extracellular trap (NET) formation. B, Quantitation of NETs. C, Representative histograms of live/dead staining of NETs. D, Quantitation of vital NETs. E, Quantitation of suicidal NETs. Data were analyzed by 2-tailed Student t test. MPO indicates myeloperoxidase.

Hypertensive Stretch Increases Neutrophil Histone H3 Citrullination and H3-Cit Exposure Leads to EC Activation, Vascular Dysfunction, and Inflammatory Cytokine Expression

Neutrophils cocultured with ECs and exposed to 5% or 10% stretch were harvested, and H3-cit was evaluated by Western blot. Neutrophils cocultured with ECs and exposed to 10% stretch exhibited increased H3-Cit compared with 5% stretch (Figure 6A and 6B). Moreover, we observed a marked increase in expression of Icam1, Vcam1, Sell, and Sele in ECs exposed to H3-Cit compared with histone H3 or vehicle (Figure 6C through 6F). Exposure of aortic rings to extracellular histone H3 or H3-cit attenuated vascular relaxation in response to acetylcholine (Figure 6G). Exposure of ECs to 10% stretch and H3-Cit significantly augmented expression of inflammatory cytokines Il6, Tgfβ, and Tnfα compared with stretched cells exposed to control (Figure 6H through 6J).

Figure 6.

Hypertensive stretch increases neutrophil histone H3 citrullination and H3-Cit exposure disrupts endothelial cell function. A, Immunoblot for H3-Cit and histone H3 on stretched neutrophils. B, Quantification of the relative expression of H3-Cit normalized to histone H3 in 5% (n=4) or 10% (n=4) stretched neutrophils. Data were analyzed by 2-tailed Mann-Whitney U test. Endothelial cells were treated with vehicle (n=4), histone H3 (n=4), or H3-Cit (n=4), and real-time PCR (RT-PCR) was used to quantify the mRNA expression of (C) Icam1, (D) Vcam1, (E) Sell, and (F) Sele normalized to the expression of Gapdh and expressed as relative expression. Data were analyzed by 1-way ANOVA with Tukey post hoc test. F, Vascular reactivity of aortic rings in response to acetylcholine (Ach) was measured following treatment with control, histone H3, and H3-Cit. Data were analyzed by a repeated-measures 2-way ANOVA with Tukey post hoc test (4 animals per group). Endothelial cells were exposed to 10% stretch and treated with vehicle (H and I, n=6; J, n=4) or H3-Cit (H–J, n=6), and RT-PCR was used to quantify mRNA expression of (H) Il6, (I) Tgfβ, and (J) Tnfα normalized to the expression of “Gapdh” and expressed as relative expression. Data were analyzed by 2-tailed Student t test (Il6 and Tgfβ) or 2-tailed Mann-Whitney U test (Tnfα).

Activation of TRPV4 Increases Neutrophil Intracellular Calcium and Increases NETosis

To determine if activation of the mechanosensitive calcium channel TRPV4 increases neutrophil intracellular calcium, isolated mouse neutrophils were treated with vehicle or 1, 10, and 100 nM of the specific TRPV4 agonist GSK1016790A. We observed a dose-dependent increase in intracellular calcium following GSK1016790A treatment measured by the mean fluorescence intensity of the calcium dye Fluo 4-AM (Figure 7A and 7B). To determine if TRPV4 agonism stimulates H3-Cit production and NETosis, we treated neutrophils with GSK1016790A and evaluated H3-Cit and NET formation by immunofluorescence. We observed a dose-dependent increase in both H3-Cit and NETosis with a marked increase at 100 nmol/L (Figure 7C through 7E).

Figure 7.

TRPV4 (transient receptor potential cation channel subfamily V member 4) activation in neutrophils augments intracellular calcium and NETosis. Murine neutrophils were treated with control (n=5), 1 nM (n=5), 10 nmol/L (n=5), and 100 nM (n=5) of the TRPV4 agonist GSK1016790A. Flow cytometry was used to detect intracellular calcium with the calcium sensitive Fluo-4 am dye. A, Representative histograms of Fluo-4 am. B, Quantitation of mean fluorescence intensity (MFI) of Fluo-4 am. Neutrophils were stained for H3-Cit, MPO (myeloperoxidase), and DNA to determine NETosis. C, Representative images of co-localized area of H3-Cit, MPO, and DNA representing neutrophil extracellular traps (NETs) in cultured neutrophils. Scale bar=100 µm. D, Quantitation of NET area per field. E, Quantitation of H3-Cit/cell count. Data were analyzed by 1-way ANOVA with Tukey post hoc test.

DISCUSSION

We have demonstrated for the first time that mice deficient in PAD4 exhibit attenuated hypertension in response to Ang II. We performed wire myography and found that Padi4^−/− mice exhibit improved EC function compared with wild-type C57Bl/6 mice following Ang II infusion. To evaluate the potential role of EC stretch on NETosis, we cocultured primary murine neutrophils with or without ECs and demonstrated that ECs augment neutrophil survival and increase NETosis. Exposure to hypertensive uniaxial stretch resulted in a significant increase in NETosis and coculture with ECs and hypertensive stretch markedly augmented NETosis. We also show that histone citrullination is increased in neutrophils exposed to hypertensive stretch. Moreover, exposure of aortic ECs to citrullinated histones leads to upregulation of the activation markers Icam1, Vcam1, Sele, and Sell. Incubation with extracellular histone H3 and H3-Cit reduces EC-dependent relaxation. Finally, we identify the role of the mechanosensitive calcium channel TRPV4 in NETosis.

Inflammation is critical for the development and maintenance of hypertension. Hypertension is associated with the production of ROS and is dependent on the function of both innate and adaptive immune cells.⁸ Neutrophils are powerful mediators of inflammation and are a significant source of ROS as well as the proinflammatory cytokines IFNγ (interferon-gamma) and TNFα all of which are causative agents in human hypertension.^36–38 In atherosclerosis, NETs induce the secretion of IL1β (interleukin 1-beta) from macrophages.³⁹ Additionally, NETs have been shown to contain IL-17 (interleukin-17).⁴⁰ Both IL1β and IL-17 have been shown to directly contribute to hypertension.^41–43 T cells require dendritic cells for activation, and both DCs and T cells infiltrate the aorta in hypertension.^8,29 Mice lacking T cells exhibit attenuated hypertension in response to Ang II.^35,44,45 The finding that cytotoxic CD8⁺ T cells are reduced in Padi4^−/− mice is intriguing. T cells have been shown to be recruited to atherosclerotic plaques following NET release.³⁹ It is conceivable that NETs similarly recruit T cells in vascular tissue in hypertension. T cells are antigen-specific and contribute to the development of hypertension. Previous work from our group has shown that isoLGs (isolevuglandins) drive both NETosis and T cell infiltration in vascular tissue.^12,29,46 Given these findings, it is conceivable that NETosis may be a source of autoantigens that drive T cell selection and tissue infiltration further contributing to hypertension. PAD4 has been shown to play a role in extracellular trap formation in other cell types including macrophages.⁴⁷ Our findings suggest that Padi4 is primarily expressed by neutrophils in hypertension, however, PAD4 may play a role in additional cell types. Hypertensive mice and humans accumulate NETs in aortic and renal tissue and, therefore, it is likely that these mechanisms contribute to NET-driven inflammation in hypertension. Direct EC injury and activation by NETs combined with the recruitment of additional inflammatory immune cell populations to vascular tissue likely contribute to hypertension.

Previous studies have shown that nitric oxide inhibits neutrophil adhesion to the endothelium whereas endothelin-1, a known driver of hypertension, augments neutrophil adhesion.^48,49 Moreover, neutrophils have been shown to adhere to and exhibit cytotoxic effects on aortic ECs from spontaneously hypertensive rats.⁵⁰ Our findings suggest that hypertensive stretch augments NETosis, and that the presence of ECs further increases NETosis. Direct mechanical factors have been shown to induce NETosis. NETs form at the site of arterial flow perturbation, suggesting a potential role of mechanical forces as a driver of NETosis in the vasculature.¹⁴ Substrate stiffness has also been shown to impact NETosis with increasing substrate stiffness leading to increased NET formation.⁵¹ This suggests that arterial stiffness in the setting of hypertension and vascular remodeling may increase the potential for NETosis. Therefore it is likely that direct mechanical stimulation mediated by neutrophil stretch contributes to NETosis. However, there remains an important potential role of the endothelium as an inducer of NETosis during hypertensive stretch. Hypertensive stretch itself activates the endothelium, which may directly stimulate NETosis.³² Gene expression analysis of ECs exposed to hypertensive stretch revealed an increase in inflammatory mediators including IL-6, IL-8, transforming TGF-β, and TNF-α.⁵² Stretch also induces the expression of adhesion molecules such as VCAM, ICAM, and integrin-α, which contribute to the binding, rolling, and transmigration of leukocytes.^52,53 It is therefore possible that secreted factors and increased adhesion contribute to NETosis. Specifically, increased adhesion of neutrophils to the endothelium likely contributes to increased mechanical forces experienced by neutrophils. We found that H3-Cit drives the expression of EC adhesion molecules. Moreover, we identified the role of H3-Cit as a driver of Il6, Tgfβ, and Tnfα expression in stretched ECs. These findings suggest that NETosis further augments EC activation and may contribute to hypertension by direct activation of inflammatory cytokine production by the endothelium.

TRPV4 is a mechanosensitive calcium channel that is highly expressed in neutrophils and contributes to ROS production, adhesion, and degranulation.^27,28 Calcium entry into neutrophils is a powerful mediator of NETosis by direct activation of PAD4 and by additional indirect mechanisms.²⁶ The finding that TRPV4 agonism induces H3-Cit accumulation suggests a direct role of TRPV4 as a driver of PAD4 activation. Moreover, an increase in NETosis following TRPV4 agonism suggests the role of TRPV4 as a mediator of stretch-induced NETosis.

The effect of NETs on ECs has been demonstrated in numerous disease states. In sepsis, NETs disrupt the EC barrier.⁵⁴ In atherosclerosis, NETs have been shown to contribute to arterial intimal injury.¹⁴ We show that H3-Cit, a product of PAD4, induces the expression of markers of EC activation. This was not observed following exposure to histone H3. Interestingly, exposure to both histone H3 and H3-Cit resulted in disruption of EC function in ex vivo vascular tissue. Extracellular histones are not specific to but are present in NETs and are proinflammatory. Histones induce tissue factor expression, nuclear translocation of NF-κB (nuclear factor kappa B), and intracellular calcium overload leading to EC dysfunction.^55,56 Consistent with these inflammatory stimuli, we found that histone H3 impaired EC-dependent vascular relaxation. However, the specificity of H3-Cit as an inducer of vascular EC activation markers is novel and suggests a unique mechanism of H3-Cit-induced EC activation. It is likely that extracellular histones and H3-Cit disrupt vascular function by unique and separate mechanisms.

The finding that hypertensive stretch augments suicidal NETosis is interesting. This may have functional consequences on EC activation whereby the complete release of cytoplasmic and nuclear contents by suicidal NETosis results in a more marked inflammatory response. Suicidal NETosis is a process that results in the death of the neutrophil.⁵⁷ This process involves the disintegration of nuclear and cytoplasmic membranes leading to the release of nuclear and granular contents.^57,58 Vital NETosis does not result in neutrophil death and is the result of nuclear budding and the release of nuclear and granular contents by exocytosis.⁵⁸ Membrane integrity is maintained in vital NETosis, and the neutrophil may continue to function as a phagocyte.^57,59 Mechanisms driving these processes have been suggested to be different, whereby vital NETosis results from toll-like receptor activation and suicidal NETosis results from numerous components including ROS production via NADPH oxidase and myeloperoxidase.^25,58 These mechanisms and processes remain poorly characterized. Hypertensive stretch leads to the production of hydrogen peroxide and additional inflammatory mediators from the endothelium.³² Moreover, hypertensive stretch reduces nitric oxide (NO) production from the endothelium.³² It is possible that these mediators may skew neutrophils toward suicidal NETosis by direct activation of NADPH oxidase or yet unexplored mechanisms. We have previously shown that isoLGs contribute to NETosis in hypertension.¹² Future studies should evaluate the role of isoLGs as drivers of TRPV4 activation, PAD4 activity, and suicidal versus vital NETosis.

The finding that ECs augment neutrophil survival is intriguing. Neutrophils are difficult to culture and survival is typically limited. We find a marked increase in neutrophil survival when cocultured with ECs for up to 36 hours. These data suggest that ECs provide support directly to neutrophils via either secreted factors or direct cell-cell interactions. This will be a topic of future studies.

In summary, we have shown that NETosis contributes to hypertension, aortic inflammation, and EC dysfunction in hypertension. Moreover, we show that EC stretch, a hallmark of hypertension, is a direct driver of NETosis. Hypertension remains one of the most important modifiable risk factors for the development of cardiovascular disease.⁶⁰ The prevalence of hypertension is increasing in the United States and blood pressure control has declined.^2,5,6 Given the important role of NETs as contributors to cardiovascular diseases, it is possible that the effects of hypertension on NETosis contribute to the development and progression of these conditions. A limitation of this study is the exclusion of female mice. Female mice are prone to the development of autoimmune conditions. Ongoing studies of NETosis and hypertension in female mice will include a study of additional systemic autoimmune manifestations. NETs are a source of autoantigens in diseases such as systemic lupus erythematosus and rheumatoid arthritis.⁶¹ It is possible that hypertension augments exposure of autoreactive adaptive immune cells to these antigens in autoimmune diseases, suggesting the importance of blood pressure control in this patient population.

ARTICLE INFORMATION

Acknowledgments

The authors thank Christian M. Warren and the flow cytometry core facility at the Nashville Veterans Affairs Hospital. Data analysis using Imaris was performed in part through the use of the Vanderbilt Cell Imaging Shared Resource. The Graphical Abstract was created with www.BioRender.com.

Author Contributions

D.M. Patrick and J. Krishnan designed the study. J. Krishnan, D.M. Patrick, N. de la Visitación, E.M. Hennen, M. Ao, and T. Ahmad performed experiments and acquired the data. J. Krishnan, T. Ahmad, M. Ao, and D.M. Patrick analyzed the data. A. Kirabo provided additional support for article preparation. All authors reviewed and revised the article. D.M. Patrick supervised the study.

Sources of Funding

This work is supported by the Veterans Affairs Biomedical Laboratory Research and Development Career Development Award (D.M. Patrick—IK2BX005376) and the National Institutes of Health (A. Kirabo—R01HL144941, R21TW012635).

Disclosures

D.M. Patrick has a patent pending regarding the use of isoLG (isolevuglandin) scavengers to treat systemic lupus erythematosus. The other authors report no conflicts.

Supplemental Material

Major Resources Table

Unedited Immunoblots

Figures S1–S6