Piezo1 in PASMCs: Critical for Hypoxia-Induced Pulmonary Hypertension Development

Abstract

BACKGROUND:

Pulmonary hypertension (PH) is a life-threatening and progressive yet incurable disease. The hallmarks of PH comprise (1) sustained contraction and (2) excessive proliferation of pulmonary arterial smooth muscle cells (PASMCs). A major stimulus to which PASMCs are exposed during PH development is altered mechanical stress, originating from increased blood pressure, changes in blood flow velocity, and a progressive stiffening of pulmonary arteries. Mechanosensitive ion channels, including Piezo1 (Piezo-type mechanosensitive ion channel component-1), perceive such mechanical stimuli and translate them into a variety of cellular responses, including contractility or proliferation. Thus, the objective of the present study was to elucidate the specific role of Piezo1 in PASMCs for PH development and progression.

METHODS:

The cell-type specific function of Piezo1 in PH was assessed in (1) PASMCs and lung tissues from patients with PH and (2) 2 mouse strains characterized by smooth muscle cell–specific, conditional Piezo1 knockout. Taking advantage of these strains, the smooth muscle cell–specific role of Piezo1 in PH development and progression was assessed in isolated, perfused, and ventilated mouse lungs, wire myography, and proliferation assays. Finally, in vivo function of smooth muscle cell–specific Piezo1 knockout was evaluated upon induction of chronic hypoxia–induced PH in these mice with insights into pulmonary vascular cell senescence.

RESULTS:

Compared with healthy controls, PASMCs from patients with PH featured an elevated Piezo1 expression and increased proliferative phenotype. Smooth muscle cell–specific Piezo1 deletion, as confirmed via quantitative real-time polymerase chain reaction and patch clamp recordings, prevented the hypoxia-induced increase in PASMC proliferation in mice. Moreover, Piezo1 knockout reduced hypoxic pulmonary vasoconstriction in isolated, perfused, and ventilated mouse lungs, endothelial-denuded pulmonary arteries, and hemodynamic measurements in vivo. Consequently, Piezo1-deficient mice were considerably protected against chronic hypoxia–induced PH development with ameliorated right heart hypertrophy and improved hemodynamic function. In addition, distal pulmonary capillaries were preserved in the Piezo1-knockout mice, associated with a lower number of senescent endothelial cells.

CONCLUSIONS:

This study provides evidence that Piezo1 expressed in PASMCs is critically involved in the pathogenesis of PH by controlling pulmonary vascular tone, arterial remodeling, and associated lung capillary rarefaction due to endothelial cell senescence.

Novelty and Significance.

What Is Known?

• Sustained vasoconstriction and structural remodeling of small pulmonary arteries are the hallmarks underlying the development and progression of all forms of pulmonary hypertension (PH).

• Lung capillary rarefaction is a major component of PH that may result from endothelial cell senescence.

• Several arguments point to mechanical stress as a major stimulus for PH development.

• The mechanosensitive ion channel Piezo1 is expressed in both pulmonary arterial endothelial and smooth muscle cells (PASMCs) and is implicated in the regulation of vascular tone and arterial remodeling.

• Although the upregulation of Piezo1 in patients with PH was recently described, the cell-specific role of Piezo1 expressed in PASMCs for PH development remains unresolved.

What New Information Does This Article Contribute?

• PASMC proliferation and arterial remodeling are decreased upon conditional smooth muscle cell–specific Piezo1 genetic deletion.

• Deletion of Piezo1 in PASMCs reduces acute and sustained hypoxic pulmonary vasoconstriction (HPV) and improves hemodynamic parameters in vivo.

• Deletion of PASMC-Piezo1 protects capillary endothelial cells from cellular senescence and prevents the loss of lung capillaries upon chronic hypoxia–induced PH.

• Smooth muscle cell-Piezo1–deficient mice are significantly protected against chronic hypoxia–induced PH development and right ventricular hypertrophy.

Our findings indicate that Piezo1 expressed in PASMCs is critically involved in controlling pulmonary vascular tone, arterial remodeling, and the onset of endothelial cell senescence during the development and progression of PH.

Meet the First Author, see p 923

Pulmonary hypertension (PH) is a life-threatening and progressive yet incurable disease, which can result in right heart hypertrophy and ultimately right heart failure.^1–4 Per definition, PH is characterized by a mean pulmonary arterial pressure above 20 mm Hg at rest⁵ and encompasses several entities, including chronic hypoxia–induced PH (CHPH), as well as pulmonary arterial hypertension.^2,5,6

While the exact causes of PH are still under investigation, it is broadly acknowledged that the hallmarks underlying the development and progression of all PH entities comprise (1) sustained vasoconstriction and (2) structural remodeling of the small pulmonary arteries, both resulting in vessel lumen obliteration and consequently an increase in vascular resistance.^2,4,7,8 In this regard, CHPH development (ie, due to high altitude or chronic lung diseases) is suggested to be dependent on persistent hypoxic pulmonary vasoconstriction (HPV), as well as an increased muscularization of pulmonary arterial vessels, chiefly as a result of the elevated proliferation of pulmonary arterial smooth muscle cells (PASMCs).^4,9 Likewise, PH occurring either as a primary disease (idiopathic or heritable) or in association with an underlying condition is usually diagnosed at the stage of coexisting vasoconstriction and marked hypertrophy of the pulmonary arteries. However, despite major advances in unraveling the pathophysiology of PH, it remains unclear whether the different forms of PH share basic mechanisms that could explain why a combination of prolonged vasoconstriction and increased proliferation of PASMCs characterizes all forms of PH.

One major stimulus to which PASMCs are exposed during PH development and progression is altered mechanical stress, originating from (1) an increased blood pressure exerting elevated strain on the vessel’s wall, (2) an altered shear stress acting on the endothelium, and (3) a progressive stiffening of the pulmonary arteries.^10,11

Thus, further investigations are needed to elucidate the molecular signaling downstream of mechanical stress as a potential common factor leading to PASMC constriction and proliferation in various forms of PH. Promising candidates for mediating mechanical forces in the vascular system are represented by mechanosensitive ion channels, such as Piezo1 (Piezo-type mechanosensitive ion channel component-1).¹² Piezo1 is a cationic nonselective stretch-activated channel characterized by a large trimeric complex with a remarkable curved shape forming an inverted dome at the cell membrane.^13–15 This dome reversibly flattens when Piezo1 opens in response to mechanical stress.^13,16 Piezo1 activation mediates calcium and potassium influx in multiple cell types¹⁷ and is also considered to play an important role during the pathogenesis of many cardiopulmonary vascular diseases, including PH.^17–19

Previous findings indicated that Piezo1 opening in endothelial cells (ECs) promotes vasorelaxation of rat pulmonary arteries (PAs) and reverses vasoconstriction in a rat model of early PH.²⁰ In addition, Piezo1 upregulation was shown to be responsible for the increased intracellular levels of free calcium in PASMCs from patients with idiopathic pulmonary arterial hypertension.¹⁸ However, it is difficult to obtain confirmatory evidence that Piezo1 expressed by PASMCs is involved in the development and progression of PH. Indeed, in pulmonary vessels, Piezo1 is expressed by both ECs^17,20 and PASMCs.¹⁸ Consequently, the effects of in vivo interventions targeting Piezo1 are difficult to interpret. Moreover, currently available pharmacological Piezo1 activators and inhibitors unfortunately only have low specificity with poor pharmacokinetic properties, which also raises challenges in interpreting some findings.²¹ Thus, although Piezo1 is anticipated to play a major role in the pathogenesis of PH, the nature of its role in PASMCs remains to be clarified.

For these reasons, the objective of the present study was to elucidate the role of Piezo1 in (1) hypoxia-induced pulmonary vasoconstriction and (2) the involvement of Piezo1 in PASMC proliferation and thus pulmonary vascular remodeling in PH. Therefore, we not only used PASMCs from patients with PH to evaluate the role of Piezo1 in PASMC proliferation but also generated 2 mouse strains characterized by smooth muscle cell (SMC)–specific, conditional Piezo1 knockout (smMHC [smooth muscle myosin heavy chain] Cre [Cre recombinase; ERT2 (estrogen ligand-binding domain)] Piezo1^lox/lox and smMHC Cre(ERT2) Piezo1^del/lox).^19,22,23 To avoid developmental effects, Piezo1 can be selectively deleted in these strains by intraperitoneal injection of tamoxifen in adult mice.^24,25 We then used these mouse strains to investigate whether SMC-Piezo1 is involved in the response to acute and chronic hypoxia in vivo and ex vivo, as well as its role in PASMC proliferation in vitro.

Methods

Data Availability

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

Detailed, expanded methods and materials are available in the Supplemental Methods and the Major Resources Table.

Animal Models and Experimental Design

Mice strains on C57BL6/J-background (Piezo1^lox/lox, Piezo1^lox/loxCre, Piezo1^del/lox, Piezo1^del/loxCre, and the Piezo1 reporter mouse line Piezo1LacZ/⁺ [gene in the lac operon that codes for the enzyme β-galactosidase]) were generated as previously described.^19,23 To generate the conditional SMC-knockdown of Piezo1 at the adult stage, the smMHC Cre(ERT2) model was used. Because smMHC is located on the Y-chromosome, only male mice could be used in our experiments to specifically delete Piezo1 in SMCs. Adult male mice were used according to institutional guidelines, which complied with national and international regulations. All animal experiments were approved by the Institutional Animal Care and Use Committee of the French National Institute of Health and Medical Research Unit 955, Créteil, France, and the German governmental authorities (RP Giessen, Hesse, Germany; Az: V 54–19 c 20 15 h 01 GI 20/10 No. G10/2017), in accordance with the German and French Animal Welfare Law and the European legislation for the protection of animals used for scientific purposes (2010/63/EU). Group size estimation was based on previous experiments performed in our laboratory or calculated using Sigma Stat 3.5 (Systat Software, Inc, San Jose, CA) with a power of 0.80 and an α of 0.05. All investigators were blinded whenever possible. Animals were kept under controlled conditions (10-hour dark/14-hour light cycle, water, and food [No. 1324, Altromin, Lange, Germany] supply ad libitum; group housed [2–5 per cage whenever possible] at 22 °C±2 °C and 55±10% humidity). Mice were handled at least for 3 days before the start of experiments to accustom them to the experimental procedures.

Human Lung Tissue Samples

Human PASMCs from patients with and without PH were provided by the UGMLC Giessen Biobank, member of the DZL Platform Biobanking. PASMCs were used from 10 donors and 12 patients with PH who had undergone lung transplantation (Table S1). Directly after explantation, PASMCs were isolated from lung tissues and snap-frozen in liquid nitrogen. The patients have given their written consent for the use of biospecimen for research purposes. The protocol was approved by the Ethics Committee of the Faculty of Medicine at the University of Giessen (Approval No. 58/15).

Immunofluorescence Studies of Human Lung Samples

For detailed information, please refer to the Methods section in the Supplemental Material.

Western Blot

For detailed information, please refer to the Methods section in the Supplemental Material.

mRNA Analysis by Quantitative Real-Time Polymerase Chain Reaction

For detailed information, please refer to the Methods section in the Supplemental Material.

LacZ Staining

For detailed information, please refer to the Methods section in the Supplemental Material.

Murine PASMC Isolation and Patch Clamp Electrophysiology

For detailed information, please refer to the Methods section in the Supplemental Material.

Hemodynamic Response of Normoxic Mice to Acute Hypoxia

For detailed information, please refer to the Methods section in the Supplemental Material.

Chronic Hypoxia Exposure of Mice

For detailed information, please refer to the Methods section in the Supplemental Material.

Assessment of PH

For detailed information, please refer to the Methods section in the Supplemental Material.

Transthoracic Echocardiography

For detailed information, please refer to the Methods section in the Supplemental Material.

Proliferation Assay

For detailed information, please refer to the Methods section in the Supplemental Material.

Wire Myography to Assess Contractility, HPV, and Stiffness of Pulmonary Arteries

For detailed information, please refer to the Methods section in the Supplemental Material.

Isolated, Ventilated, and Perfused Lung Experiments

For detailed information, please refer to the Methods section in the Supplemental Material.

Immunofluorescence Studies of Mouse Lung Samples

For detailed information, please refer to the Methods section in the Supplemental Material.

Immunohistochemical Studies of Mouse Lung Samples

For detailed information, please refer to the Methods section in the Supplemental Material.

Statistical Analysis

Data were presented as mean±SEM. The paired or unpaired Student t tests were performed for comparing the 2 groups. Familywise error rates were controlled by the Dunnett, Sidak or uncorrected Fisher least significant difference multiple comparison procedures. For groups with small sample sizes, in which normality cannot be reliably established (as determined by a D’Agostino and Pearson omnibus normality test for Gaussian distributions), a nonparametric test (Mann-Whitney U test) was used to determine statistical differences. The calculations were done by using GraphPad Prism 8. A P value of <0.05 was considered statistically significant.

Results

Piezo1 Expression in PASMCs Is Associated With Remodeling of Pulmonary Vessels in Patients With Idiopathic Pulmonary Arterial Hypertension

We first examined Piezo1 localization and expression in paraffin-embedded lung slices from patients suffering from idiopathic pulmonary arterial hypertension and healthy controls (Figure 1A) that were matched in age and sex (Table S1). As visualized by confocal immunofluorescence microscopy, Piezo1 expression was found to be upregulated in the medial layer of thickened small pulmonary arteries from patients with idiopathic pulmonary arterial hypertension compared with controls (Figure 1A). Furthermore, coimmunostaining against α-SMA (α-smooth muscle actin, a marker for SMCs) and elastin (as a marker for connective tissues) confirmed Piezo1 localization in the smooth muscle layer of pulmonary arteries. This finding was further confirmed in Western Blot experiments showing a markedly increased Piezo1 expression in primary PASMCs from patients with PH versus healthy controls when normalized to β-actin or α-SMA expression (Figure 1B and 1D). To question whether Piezo1 overexpression in PASMCs was shared by other lung pathologies, we additionally performed immunofluorescence studies of remodeled pulmonary vessels from patients with chronic obstructive pulmonary disease. As shown in Figure S1, a strong Piezo1-immunostaining colocalized with α-SMA at sites of pulmonary remodeling was also found in patients with chronic obstructive pulmonary disease.

Figure 1.

Elevated Piezo1 expression in the medial layer of pulmonary vessels in patients with idiopathic pulmonary arterial hypertension (IPAH). A, Representative photomicrographs of confocal immunofluorescence staining of Piezo1 (white), α-SMA (α-smooth muscle actin; red), elastin (green), and 4′,6-diamidino-2-phenylindole (DAPI; blue) in lung tissue of 2 patients with IPAH and respective controls from healthy donors. Remodeled pulmonary vessels of patients with IPAH depict an enhanced expression of Piezo1 in the medial layer of pulmonary vessels, as confirmed by double-staining against SMA. Scale bars, 50 µm. B, Western blot analysis and (C and D) densitometry of Piezo1 expression in pulmonary arterial smooth muscle cells (PASMCs) from pulmonary hypertension (PH; n=6) and control patients (n=6). Differences were assessed via the unpaired t test and displayed as P value.

Generation and Validation of SMC-Specific Conditional Piezo1-Knockout Mice

Taking advantage of a LacZ Piezo1 reporter mouse,¹⁹ we observed a strong expression of Piezo1 in the media of pulmonary arteries (Figure 2A). To further evaluate the specific role of Piezo1 in PASMCs for PH development and progression, conditional SMC-specific Piezo1-knockout mouse lines were used¹⁹ (Figure S2A). Therefore, 2 mouse strains, Piezo1^lox/lox and Piezo1^del/lox, were each bred with a transgenic mouse line that expresses CreERT2 in SMCs.²⁶ The expression of this fusion protein is driven by the smMHC promoter to ablate Piezo1 specifically in SMCs and only when induced by tamoxifen (ERT2 ligand) injections. Of note, 1 Piezo1 allele is constitutively deleted in the Piezo1^del/lox strain, thus allowing optimal knockdown of Piezo1 upon Cre-activation and the resulting deletion of the second floxed Piezo1 allele.^19,22,23 Piezo1 knockdown efficacy was then evaluated via quantitative real-time polymerase chain reaction in isolated and endothelial-denuded pulmonary arteries. As shown in Figure 2B, Piezo1 mRNA was considerably lower in all 3 knockout strains. As anticipated, the highest knockdown efficacy was observed in Piezo1^del/loxCre mice, without any compensatory changes in Piezo2 expression levels (Figure S2B). However, body weight remained unaltered upon SMC-Piezo1 deletion (Figure 2C). Notably, we did not find any modification in cardiac systolic function as assessed by left ventricular ejection fraction and left ventricular fractional shortening and in regional myocardial contractility as measured by strain rate in mutant mice versus controls, with similar heart rate (Table S2).

Figure 2.

Piezo1 expression and validation of knockdown in pulmonary arterial smooth muscle cells (PASMCs). A, Sections of pulmonary arteries (PAs) from control Piezo1^+/+ and reporter Piezo1^LacZ/+ mice demonstrating expression of Piezo1 (blue staining) in the smooth muscle layer of PAs. B, Quantitative real-time polymerase chain reaction (qRT-PCR) data illustrating the progressive knockdown of Piezo1 in PAs of Piezo1^lox/lox (n=4), Piezo1^lox/loxCre (n=3), Piezo1^del/lox (n=3), and Piezo1 ^del/loxCre (n=4) mice. C, No change in the body weight upon Piezo1 deletion was observed (Piezo1^lox/lox [n=5], Piezo1^lox/loxCre [n=6], Piezo1 ^del/lox [n=5], and Piezo1 ^del/loxCre [n=4]). D, Cell-attached patch clamp recordings of Piezo1 currents in response to fast pressure pulse stimulation, recorded at a holding potential of −80 mV (top) and at the predicted reversal potential for Piezo1 (0 mV; bottom). E, Pressure-effect curves for Piezo1 current amplitude in Piezo1^lox/lox (n=13) and Piezo1^del/loxCre mice (n=18). Differences were assessed via the Mann-Whitney U test and displayed as P values.

To functionally assess Piezo1 deletion, we recorded stretch-activated channels in freshly isolated PASMCs in the cell-attached patch clamp configuration (Figure 2D and 2E). Increasing negative pressure pulses were applied every 10 seconds at holding potentials of either −80 or 0 mV. As shown in Figure 2D, pressure-dependent stretch-activated channel currents were only observed for Piezo1^lox/lox PASMCs at a holding potential of −80 mV while being absent at 0 mV (the predicted reversal potential for Piezo1^19,22,23). Compared with PASMCs from control mice (Piezo1^lox/lox), stretch-activated channel currents in response to negative pressure pulses were greatly reduced upon Piezo1 knockdown (Piezo1^del/lox Cre; Figure 2D, right, and E).

Piezo1 Deletion Impairs Contractility in Isolated Pulmonary Arteries

To determine the role of PASMC-Piezo1 in general vasoreactivity and HPV in particular, wire myography was performed on freshly isolated and endothelial-denuded intrapulmonary arteries (IPAs) of all 4 strains (Figure 3; Figure S3). According to the scheme shown in Figure 3A, endothelial-denuded IPAs were first exposed to a wake-up protocol including exposure to high potassium-enriched physiological saline solution and phenylephrine to ascertain general vasoreactivity, as well as acetylcholine to provide evidence for successful removal of the endothelium (Figure S3B). Afterward, IPAs were subjected to hypoxic conditions to determine HPV. Finally, passive properties, that is, arterial stiffness, were measured in calcium-free conditions.

Figure 3.

Piezo1 deficiency diminishes the contractility of small intrapulmonary arteries. A, Schematic representation of the procedure of wire myography experiments on endothelial-denuded intrapulmonary arteries (IPAs) of Piezo1-deficient mice and their respective controls. Arteries were first exposed to a wake-up protocol to confirm the viability and assess the general contractility of the vessels. Therefore, the arteries were stimulated 3-fold with exposure to potassium-enriched physiological saline solution (KPSS), separated by wash-out steps. As the final part of the wake-up procedure, phenylephrine (PE, 10 µmol/L) was added to the bath solution. The removal of the endothelium was confirmed by application of acetylcholine (ACh, 10 µmol/L). Only those arteries that exhibit no ACh-mediated vasodilation were regarded as endothelium-denuded and used for further experiments. Hypoxic pulmonary vasoconstriction (HPV) was initiated after pretone with 15-mmol/L potassium chloride (KCl). Finally, arterial stiffness was determined under calcium (Ca²⁺)-free conditions. B, HPV in endothelial-denuded IPAs of Piezo1-deficient mice (Piezo1^del/loxCre, purple; n=11) and their respective controls (Piezo1^lox/lox, gray; n=12). Data are normalized to the tension value before the onset of the hypoxic challenge and displayed as mean±SEM (dotted lines). HPV was determined by calculating the area under the curve (AUC, displayed as a shaded area). C, Statistical analysis of the experiments depicted in B (Piezo1^lox/lox [n=12], Piezo1^lox/loxCre [n=9], Piezo1^del/lox [n=9], and Piezo1^del/loxCre [n=11]). D, General contractility of the IPAs in response to KPSS and (E) PE was reduced in Piezo1-deficient mice (Piezo1^lox/lox [n=12], Piezo1^lox/loxCre [n=9], Piezo1^del/lox [n=9], and Piezo1^del/loxCre [n=11]). F, Stiffness of IPAs remained unaltered by Piezo1 deficiency (Piezo1^lox/lox [n=12], Piezo1^lox/loxCre [n=9], Piezo1^del/lox [n=9], and Piezo1^del/loxCre [n=11]). Data were statistically analyzed by the 1-way ANOVA and the uncorrected Fisher least significant difference test for multiple comparisons and displayed as P values. Only differences regarded as significant changes (P<0.05) are displayed as P values.

In line with previous reports,^27,28 hypoxia induced a monophasic, gradually developing sustained contraction in IPAs of control Piezo1^lox/lox mice (Figure 3B), which was largely diminished in SMC-specific Piezo1-knockout lines (Figure 3B and 3C). A similar picture emerged when assessing the general contractility in IPAs. The exposure to the potassium-enriched physiological saline solution (Figure 3D) and phenylephrine (Figure 3E) induced a pronounced elevation in the tension of Piezo1^lox/lox-IPAs that were significantly diminished upon Piezo1 deletion in PASMCs (Figure 3D and 3E). However, the passive properties of IPAs, as determined by stiffness measurements, were unaltered between the 4 genotypes (Figure 3F).

PASMC-Piezo1 Is Involved in HPV In Vivo and in Isolated, Perfused, and Ventilated Mouse Lungs

Considering the critical role of PASMC-Piezo1 for HPV in IPAs, we next investigated the physiological consequences of PASMC-Piezo1 deletion for acute HPV in vivo. Therefore, anesthetized mice were ventilated with room air before being acutely exposed to a hypoxic gas mixture, while the right ventricular systolic pressure (RVSP) was simultaneously recorded. In Piezo1^lox/lox controls, acute hypoxia increased RVSP by 6.5±1.2 mm Hg (Figure 4A and 4B). In contrast, the hypoxia-induced rise in RVSP was considerably lower in PASMC-Piezo1–deficient mice with an elevation of 1.7±0.5 mm Hg in Piezo1^del/loxCre mice (Figure 4B). To additionally examine, whether PASMC-Piezo1 deletion not only affects acute (peaking within 10 minutes) but also sustained HPV (observed over a time period of 180 minutes), further experiments were performed on isolated, constant-flow-perfused, and ventilated mouse lungs (Figure 4C and 4F). While the initial pulmonary arterial pressure was similar between the distinct genotypes (Figure 4C), both acute (Figure 4D and 4E) and sustained HPVs (Figure 4D and 4F) were largely reduced in lungs from Piezo1^del/loxCre mice compared with controls. Quantitative real-time polymerase chain reaction confirmed knockdown of Piezo1 in lung homogenates (of note Piezo1 is not deleted in the lung parenchyma upon smMHCCre activation; Figure 4G).

Figure 4.

Piezo1 deficiency reduces hypoxic pulmonary vasoconstriction. A, Statistical analysis of changes in right ventricular systolic pressure (RVSP) before (−) and 5 minutes after hypoxic ventilation with 8% O₂ (+, shaded in light purple; Piezo1^lox/lox [n=6], Piezo1^lox/loxCre [n=9], Piezo1^del/lox [n=17], and Piezo1^del/loxCre [n=11]). Data were statistically analyzed by paired Student t test. B, Pressure differences in RVSP as calculated by ∆RVSP=RVSP^hypoxia−RVSP^normoxia revealed a blunted hypoxic pulmonary vasoconstriction (HPV) response in Piezo1-deficient mice (Piezo1^lox/lox [n=6], Piezo1^lox/loxCre [n=9], Piezo1^del/lox [n=17], and Piezo1^del/loxCre [n=11]). Data were statistically analyzed by the 1-way ANOVA and the uncorrected Fisher least significant difference test and displayed as P values. C, Baseline pulmonary arterial pressure (PAP) in isolated, ventilated, and perfused mouse lungs (Piezo1^lox/lox [n=5], Piezo1^lox/loxCre [n=6], Piezo1^del/lox [n=5], and Piezo1^del/loxCre [n=4]) before the onset of hypoxic ventilation is not affected by Piezo1 deficiency. D, Time course of the strength of HPV during 180 minutes of hypoxic ventilation (1% O₂) in isolated, ventilated, and perfused mouse lungs. PAP was normalized to the value before the onset of the hypoxic challenge (∆PAP) and displayed for lungs from control (Piezo1^lox/lox, gray; n=5) and Piezo1-deficient mice (Piezo1^del/lox, purple; n=4). Strength of HPV was calculated for both, the acute phase (peak within 5–10 minutes) and the sustained phase (from 30 to 180 minutes) by determining the area under the curve (AUC, shaded area). E, Statistical analysis for acute and (F) sustained phase of HPV. G, Piezo1 deficiency was confirmed via quantitative polymerase chain reaction (qPCR); n=8 per genotype. Data were statistically analyzed by the 1-way ANOVA and the uncorrected Fisher least significant difference test. Only differences regarded as significant changes (P<0.05) are displayed as P values.

Piezo1 Knockout in PASMCs Confers a Considerable Protection Against Chronic Hypoxia–Induced PH in Mice

Our finding of Piezo1 being crucial for both acute and sustained HPVs prompted us to test whether Piezo1 gene deletion in PASMCs could prevent CHPH development in mice. After hypoxia exposure for 21 days, RVSP was significantly lower, and right ventricular hypertrophy was less severe in Piezo1-deficient mice than in control mice (Figure 5A and 5B). Furthermore, distal pulmonary vessels exhibited less muscularization in Piezo1-knockout mice than in the respective controls (Figure 5C and 5D), with a smaller percentage of dividing PCNA (proliferating cell nuclear antigen)–positive pulmonary vascular cells after chronic hypoxic exposure (Figure 5E and 5F).

Figure 5.

Piezo1 knockout in pulmonary arterial smooth muscle cells (PASMCs) partially confers protection against chronic hypoxia–induced pulmonary hypertension (PH) in mice. A, Right ventricular systolic pressure (RVSP) of mice after 3 weeks of normoxic (white area) or hypoxic (shaded area) exposure (normoxia:

Piezo1^lox/lox [n=3], Piezo1^lox/loxCre [n=7], Piezo1^del/lox [n=5], and Piezo1^del/loxCre [n=7]; hypoxia: Piezo1^lox/lox [n=6], Piezo1^lox/loxCre [n=6], Piezo1^del/lox [n=6], and Piezo1^del/loxCre [n=6]). B, Right heart hypertrophy, depicted as the Fulton index (right ventricle [RV]/[left ventricle (LV)+septum (S)]; normoxia: Piezo1^lox/lox [n=5], Piezo1^lox/loxCre [n=5], Piezo1^del/lox [n=5], and Piezo1^del/loxCre [n=5]; hypoxia: Piezo1^lox/lox [n=6], Piezo1^lox/loxCre [n=6], Piezo1^del/lox [n=6], and Piezo1^del/loxCre [n=6]). C, Representative photographs and (D) values of mean muscularization of small (20–70 µm) pulmonary vessels from paraffin-embedded lung sections (normoxia: Piezo1^lox/lox [n=5], Piezo1^lox/loxCre [n=5], Piezo1^del/lox [n=5], and Piezo1^del/loxCre [n=6]; hypoxia: Piezo1^lox/lox [n=6], Piezo1^lox/loxCre [n=6], Piezo1^del/lox [n=6], and Piezo1^del/loxCre [n=6]). Scale bars, 50 µm. E, Representative immunofluorescence staining against PCNA (proliferating cell nuclear antigen, pink) and (F) statistical analysis for evaluation of proliferation in normoxic and chronic hypoxic mice (normoxia: Piezo1^lox/lox [n=5], Piezo1^lox/loxCre [n=5], Piezo1^del/lox [n=5], and Piezo1^del/loxCre [n=5]; hypoxia: Piezo1^lox/lox [n=5], Piezo1^lox/loxCre [n=5], Piezo1^del/lox [n=5], and Piezo1^del/loxCre [n=6]). Scale bars, 50 µm. Data were statistically analyzed by the 1-way ANOVA and the uncorrected Fisher least significant difference test. Only differences regarded as significant (P<0.05) are displayed as P values.

Piezo1 Is Involved in Hypoxia-Induced PASMC Proliferation in Mice

To assess the potential role of Piezo1 in mediating PASMC proliferation in different conditions, we examined the growth response of PASMCs isolated from mice that have been exposed to either normoxia or chronic hypoxia (Figure 6A). In accordance with our previous results demonstrating that PASMCs from chronically hypoxic mice show an increased proliferative phenotype (Figure 5E and 5F), we found that PASMCs from chronically hypoxic mice grew faster than those from their normoxic counterparts in the presence of 0.2% or 10% serum (Figure 6A). This higher growth rate of cells from chronically hypoxic mice compared with their wild-type controls was abolished in PASMCs from Piezo1-deficient mice (Figure 6A). By using knockout strains that were based on an SMC-specific promoter, any influence through other mesenchymal cells, for example, fibroblasts and myofibroblasts, can be excluded.

Figure 6.

Modulation of Piezo1 activity affects the proliferation of pulmonary arterial smooth muscle cells (PASMCs) from hypoxic mice and patients with pulmonary hypertension (PH). A, Proliferation assays of PASMCs isolated from control (Piezo1^lox/lox: normoxia [n=6] and hypoxia [n=5]) and Piezo1-deficient mice (Piezo1^lox/loxCre: normoxia [n=4], hypoxia [n=7]; Piezo1^del/loxCre: normoxia [n=8] and hypoxia [n=5]) exposed to normoxia or to chronic hypoxia (21 days). Proliferation was assessed in the presence of 0.2 and 10% FCS (fetal calf serum) and after exposure to increasing concentrations of Yoda1 (selective activator of the Piezo1 ion channel) and (B) GsMTx4. Values are expressed as percent of control values in starvation DMEM (Piezo1^lox/lox: n=8–9; Piezo1^del/loxCre: n=6–8). Data are shown as mean±SEM from at least 3 different experiments. C, Effect of increasing concentrations of Yoda1, (D) margaric acid (MA), and (E) GsMTx4 (GsM toxin 4, a peptide toxin) on the proliferation of PASMCs from patients with PH (n=6–8) and controls (n=8–9) in the presence of 0.2 and 2% FCS. Values are expressed as percent of control values in starvation Dulbecco's Modified Eagle Medium (DMEM). Data are mean±SEM of 8 to 10 values. Data were statistically analyzed by the 2-way ANOVA and the uncorrected Fisher least significant difference test. P values are indicated for comparison among group means.

In addition, PASMCs from the different mouse strains were subjected to increasing doses of the Piezo1-activator Yoda1 (selective activator of the Piezo1 ion channel; Figure 6A) and the Piezo1-inhibitor GsMTx4 (GsM toxin 4, a peptide toxin; Figure 6B). A stimulatory and dose-dependent effect of Yoda1 on PASMC growth was observed in cells from control mice in the presence of 10% FCS (fetal calf serum; Figure 6A). In contrast, cultured PASMCs from either Piezo1^lox/loxCre or Piezo1^del/loxCre mice did not exhibit any stimulatory effect in response to Yoda1 (10% serum; Figure 6A). Reciprocally, the inhibitory effect of GsMTx4 on PASMC growth was more pronounced in cells from control mice stimulated by serum (10% FCS) and largely diminished in cells from Piezo1^del/loxCre mice (Figure 6B).

PASMCs From PH Patients Feature an Increased Proliferative Phenotype

Based on these results obtained in mouse PASMCs, we studied the effects of Yoda1 on the growth rate of PASMCs from patients with PH compared with those of healthy donors (Figure 6C). The PH-PASMCs, expressing higher levels of Piezo1 than PASMCs from healthy donors (Figure 1C), also demonstrated a higher proliferating capacity (Figure 6C and 6E) that was diminished upon treatment with the antagonists margaric acid²⁹ (Figure 6D) and GsMTx4³⁰ (Figure 6E), which both caused a net decrease in the proliferation rate (Figure 6D and 6E). Again, a stimulatory effect of Yoda1 on PASMC growth was observed, which was blunted in cells isolated from control individuals (Figure 6C). In line, small interfering RNA–mediated silencing of Piezo1 in PH-PASMCs decreased proliferation (Figure S4).

Piezo1 Activity Is Unaffected by Acute Hypoxia

To elucidate the mechanism behind Piezo1-mediated hypoxia-induced increases in contraction and proliferation, we performed patch clamp recordings to explore the potential direct impact of acute hypoxia on Piezo1 activity. Therefore, HEK293T (Human Embryonic Kidney 293 cells, transformed) cells overexpressing Piezo1 were exposed to acute hypoxia, while ionic currents were recorded in whole-cell configuration. As depicted in Figure S5, Piezo1-mediated current was unaffected by acute hypoxia. To confirm Piezo1 expression, the cells were exposed to Yoda1 following hypoxic exposure. As expected, Yoda1 induced a significant increase in Piezo1 activity that was inhibited by subsequent application of GsMTx4.

Deletion of Piezo1 in SMCs Prevents Microvascular Cell Senescence Under Chronic Hypoxia

Because of the marked effect of Piezo1 on hypoxic vasoconstriction and pulmonary vascular remodeling, we further explored whether Piezo1 deficiency could alter pulmonary vascular cell senescence, a major component of PH.^31,32 Indeed, Piezo1 was reported to affect cellular senescence in a positive or negative manner in different tissues by modifying the cellular response to mechanical stress.^33,34 To address this issue, we analyzed the expression of senescence markers in lung tissues obtained from Piezo1^lox/lox- and Piezo1-deleted mice that were previously subjected to chronic hypoxia. As expected, we found an accumulation of senescent cells stained for p16 and p21 (2 reliable markers of cellular senescence) in Piezo1^lox/lox mice exposed to chronic hypoxia (Figure 7A and 7D). P16 staining and p21 staining were localized in both remodeled vessels and the distal lung (Figure 7A and 7C; Figure S6). In contrast, p16 immunostaining and p21 immunostaining were low in chronically hypoxic Piezo1^del/loxCre mice (Figure 7A and 7D; Figure S6). Expression of p16 protein was found in both PASMCs and microvascular ECs as shown by double-immunofluorescence staining for p16 and α-SMA or CD31 (Cluster of Differentiation 31), respectively (Figure S6A and S6B). However, the reduction of p16-stained PASMCs in Piezo1-deleted mice was difficult to quantify, given the low degree of remodeling of their vessel wall and the difficulty in reliably assessing the number of PASMCs costained for p16 and α-SMA (Figure S6B). In contrast, chronically hypoxic Piezo1^del/loxCre mice compared with Piezo1^lox/lox mice exhibited a marked reduction in p16-positive capillary ECs costained for CD31 (Figure S6B). Consequently, microvascular EC density, as assessed with CD31 labeling of the distal lung, was much higher in Piezo1^del/loxCre mice than in Piezo1^lox/lox mice exposed to chronic hypoxia (Figure 7E and 7F; Figure 6B). Furthermore, we observed that the DNA damage markers 53BP1 (p53 Binding Protein 1), 8-hydroxy-2’-deoxyguanosine, and γ-H2AX (Phosphorylated Histone H2AX at Serine 139) markedly increased in chronically hypoxic Piezo1^lox/lox mice but not in chronically hypoxic Piezo1^del/loxCre mice (Figure S7A through S7F). Our interpretation of these findings is that the combination of reduced pulmonary vasoconstriction and vascular remodeling enabled by Piezo1 deficiency protected distal vessels from oxidative stress–induced EC senescence and, thereby, from the rarefaction of the capillary network.

Figure 7.

Piezo1 deletion in vascular smooth muscle cells protects Piezo1^del/loxCre mice against hypoxia-induced endothelial cell senescence and microvascular rarefaction. A, Representative pictures showing immunofluorescence staining of p16 (white) in normoxic and chronic hypoxic murine lungs. The zoomed areas are indicated by rectangles. Blue: 4′,6-diamidino-2-phenylindole (DAPI) nuclear staining; green: elastin autofluorescence. Scale bars, 200 µm. B, Statistical analysis representing the percentage of p16+ cells in Piezo1^lox/lox (normoxia: n=7; hypoxia: n=4) and Piezo1^del/loxCre mice (normoxia: n=5; hypoxia: n=6). C, Representative micrographs showing immunofluorescence staining of p21 (brown) in small pulmonary vessels (left) and lung parenchyma (right). Blue: hematoxylin nuclear staining. Scale bars, 50 µm. D, Statistical analysis representing the percentage of p21+ lung cells in Piezo1^lox/lox (normoxia: n=7; hypoxia: n=3) and Piezo1^del/loxCre mice (n=4). E, Representative micrographs showing immunofluorescence staining of CD31 (red, a marker of endothelial cells). Blue: DAPI nuclear staining. Scale bars, 200 µm. F, Statistical analysis representing the percentage of CD31⁺ area in lung tissues of Piezo1^lox/lox (normoxia: n=5; hypoxia: n=4) and Piezo1^del/loxCre mice (normoxia: n=4; hypoxia: n=5). B, D, and F, Data are expressed as individual values per mice and mean±SEM for groups of mice. Data were statistically analyzed by the 2-way ANOVA and the uncorrected Fisher least significant difference test. Only differences regarded as significant changes (P<0.05) are displayed as P values.

Discussion

In the present study, we investigated Piezo1 function in human and murine PH,¹⁷ with a special focus on the role of Piezo1 in PASMCs. Our study identifies Piezo1 as a major determinant of both the regulation of contractility and proliferation of PASMCs during PH development and progression. By combining studies of PASMCs from patients with PH and control patients, as well as conditional SMC-specific Piezo1-knockout mouse models, we demonstrated that Piezo1 expressed by PASMCs is a key modulator of pulmonary vessel tone and remodeling. Moreover, Piezo1 was found to be crucial to the extent of HPV in vivo and enhanced various constricting and proliferative stimuli exerted on PASMCs during PH development and progression. In addition, Piezo1 deficiency in PASMCs contributes to the maintenance of distal pulmonary capillaries, associated with the reduction in the number of senescent microvascular ECs through inhibition of HPV and pulmonary vascular remodeling.

Piezo1 is a mechanosensitive nonselective cationic channel that is expressed in blood vessels by both ECs and SMCs.^19,21 It is activated by increased blood flow (laminar shear stress or turbulent flow),^35,36 as well as by cell membrane stretch related to blood pressure elevation.¹⁹ Piezo1 activation by steady shear stress in ECs causes vasorelaxation (flow-mediated dilation) and is implicated in vascular inflammation caused by turbulent flow.^21,37 Given the high flow rate in pulmonary vessels and their marked distensibility,³⁸ combined with the intrinsic ability of PASMCs to constrict in response to hypoxia (thereby increasing cell membrane tension in responsive vessels⁹), the role of Piezo1 in controlling the pulmonary vascular bed is expected to be significant. Previous studies using mice IPAs demonstrated that endothelial Piezo1 contributes to pulmonary vasodilation and protects against hypoxia-induced vasoconstriction.³⁹ In addition, Piezo1 activation in rat pulmonary arterial ECs by Yoda1 causes vasorelaxation and reverses vasoconstriction in a rat model of early PH.^18,20 However, Piezo1 deletion from pulmonary arterial ECs had no effect on PH development in experimental animals,^17,39 whereas systemic injection of a Piezo1 channel blocker was shown to ameliorate chronic hypoxia–induced PH in wild-type mice.¹⁷ Piezo1-mediated increase in intracellular calcium was anticipated to be associated with enhanced human PASMC contraction and proliferation.¹⁸ Notably, upregulated expression and increased activity of Piezo1 have been observed in PASMCs from patients with PH.¹⁸ Moreover, Yoda1 causes vasoconstriction of endothelium-denuded rat PAs.^18,20 Although these observations suggested an important and selective homeostatic role for Piezo1 in PAs, they provided no answer about whether Piezo1 may be involved in PH development and progression.

First, we identified the cells exhibiting the highest Piezo1 expression levels in established PH. In accordance with data reported by Liao et al,¹⁸ we found stronger immunostaining of the media of remodeled vessels from patients with PH compared with controls. Interestingly, findings were similar in remodeled vessels from patients with chronic obstructive pulmonary disease, indicating that Piezo1 overexpression in PASMCs was common in different lung pathologies. Of note, Piezo1 overexpression persisted in cultured PASMCs collected from PAs of patients with PH. These findings led us to focus on the specific role of Piezo1 expressed by PASMCs in PH development and progression.

The generation of conditional SMC-specific Piezo1-knockout mice gave us the unique opportunity to study the physiological role of Piezo1 in adult mice. Piezo1 mRNA assays of IPAs isolated from these mice demonstrated that Piezo1 expression levels differed across our mouse lines, as previously observed in the systemic circulation, as well as in renal tubular cells.^19,22,23 In isolated and endothelial-denuded small IPAs from Piezo1^lox/loxCre and Piezo1^del/loxCre mice, the contractile response to cell depolarization (potassium chloride application) or phenylephrine, as well as the response to acute hypoxia, was markedly reduced. By contrast, Piezo1 knockout in SMCs of resistance arteries within the systemic circulation (caudal artery or basilar artery) does not affect contractility (in response to potassium chloride–induced depolarization or phenylephrine stimulation), as well as the myogenic tone in response to increased luminar pressure.²³ Thus, SMC-Piezo1 plays a specific role in the regulation of PA contractility. In contrast, Piezo1 deletion had no effect on arterial stiffness as assessed by measuring vessel diameter changes in response to increasing strain. Of note, because these experiments were performed in IPAs without endothelium, these results clearly indicate that HPV, at least in this ex vivo model, depends on Piezo1 activity in SMCs and, thus, is completely independent of ECs. Next, we investigated whether similar findings were obtained in more integrated experimental models. In intact animals, the pulmonary pressure response to acute hypoxia was markedly blunted in both Piezo1^lox/loxCre and Piezo1^del/loxCre animals and was intermediate in Piezo1^del/lox mice. To further investigate the influence of Piezo1 on the intrinsic properties of the pulmonary circulation without interference from systemic effects, we used the isolated, ventilated, and perfused mouse-lung model to investigate the pressure response to hypoxia at constant flow. As previously described, this model allows the distinction between 2 specific phases of contraction, namely, acute (onset within 10 minutes) and sustained HPV (observed over a time period of 180 minutes). In our mouse lines, Piezo1 deficiency diminished both phases of HPV, the early transient peak, and the sustained elevation of pulmonary pressure in response to hypoxia. In conclusion, our data on pulmonary vasoreactivity obtained via different experimental approaches clearly indicate that Piezo1 is a key modulator of the contractile response of pulmonary vessels, thereby influencing HPV.

Because sustained pulmonary vessel constriction is viewed as a major mechanism underlying the development of PH in response to prolonged hypoxia exposure,⁹ we further investigated the development of CHPH in our mouse lines. A key finding is that Piezo1-depleted mice were protected against CHPH development, with PH severity being roughly commensurate with the Piezo1 expression level in our various mutant mice. These results constitute strong evidence that Piezo1 has a major impact on the development of CHPH. Of note, no difference was seen during normoxia exposure, suggesting that the role of Piezo1 in controlling pulmonary vessels may become more important during the development of CHPH.

The hemodynamic parameters determined in chronically hypoxic mice were measured at least 2 hours after the return to normoxia. Thus, the effect of Piezo1 on pulmonary vasoreactivity may no longer have been responsible for the differences across mouse strains at the time of the phenotypic assessment. Consistent with this possibility, Piezo1-mutant mice exhibited a reduction in pulmonary vascular remodeling commensurate with the reductions in RVSP and the Fulton index, again with good proportionality to the magnitude of Piezo1 knockout in PASMCs. We, therefore, investigated whether Piezo1 deletion affected the proliferation of PASMCs from both normoxic and chronically hypoxic mice. As expected, PASMCs from chronically hypoxic mice grew faster than PASMCs from normoxic mice, despite being exposed to identical culture conditions with various concentrations of FCS.⁴⁰ Notably, the growth rate, which was increased for cells from chronically hypoxic mice, was far slower for cells from Piezo1-depleted mutant mice under both FCS concentrations. Taken together, these results indicate a strong link between the proliferative phenotype of PASMCs and the level of Piezo1 expression under PH conditions. To better assess whether this proliferative phenotype was altered by modulating Piezo1 activity, we exposed cells to Piezo1 activators and antagonists previously used in pharmacological studies, that is, 1 short saturated lipid (margaric acid) and 1 spider venom peptide (GsMTx4). Although the specificity of these compounds is questionable and may depend on the experimental conditions used,²¹ their effects on cell proliferation were absent in cells from Piezo1-deficient mice compared with cells from control mice. In particular, the inhibitory effects of GsMTx4 were more marked in PASMCs from Piezo1^lox/lox mice than from Piezo1-knockout mice, further supporting a contribution of Piezo1 to the increased proliferative phenotype of PASMCs in the setting of PH.

These results obtained using mouse PASMCs prompted us to study the effects of Yoda1, GsMTx4, and margaric acid on the growth rate of PASMCs from patients with PH versus healthy controls. Our own findings and others demonstrated that PASMCs from patients with PH express higher levels of Piezo1 compared with controls.²¹ In addition, we confirmed the proproliferative phenotype of PH-PASMCs that have been previously reported.⁴¹ However, the inhibitory effect of the Piezo1 antagonist GsMTx4 on PASMC growth did not differ between PASMCs from patients with PH versus controls. Interestingly, feeding cells with margaric acid that was previously reported to inhibit Piezo1 via influencing the cell membrane’s lipid composition²⁹ reduced the proliferation of PASMCs from patients with PH. Moreover, the high proliferating capacity of PASMCs from PH patients was blunted upon small interfering RNA–mediated knockdown of Piezo1. Overall, these results support a major functional role for Piezo1 in controlling pulmonary vessel tone and remodeling by acting on both the general contractility and the proliferation of PASMCs. Reciprocally, decreasing Piezo1 activity slowed the development of PH, notably via a marked decrease in the extent of HPV. This is a major mechanism in group 3 PH, but most PH types result from the combination of sustained vasoconstriction and elevated proliferation.

In recent studies, we showed that 1 major component of PH development is the onset of cellular senescence that affects PASMCs in remodeled vessels and capillary ECs, resulting in capillary rarefaction.^31,32 We, therefore, explored the hypothesis that the improvement in pulmonary hemodynamics induced by Piezo1 deficiency might be linked to some effects on pulmonary vascular cell senescence. Indeed, Piezo1 has been shown to affect cellular senescence in a positive or negative manner in different tissues by modifying the cellular response to mechanical stress.^33,34 Notably, Piezo1 deficiency in PASMCs protected against senescence of microvascular ECs and, consequently, against the rarefaction of the pulmonary capillary network that is closely associated with PH development.

Possible mechanisms by which Piezo1 deletion in SMCs protected against ECs senescence are likely to be indirect. The potential contribution of Piezo1-dependent depolarization transmitted from SMCs to ECs is an attractive hypothesis⁴² and a possible contribution of mechanical stress associated with the increase in PA pressure. Indeed, the combination of reduced pulmonary vasoconstriction and vascular remodeling enabled by Piezo1 deficiency protected distal vessels from increased pulmonary arterial pressure and possibly mechanical stress–induced EC senescence. Consistent with the latter hypothesis, mechanical stress applied to vascular cells has been shown to result in oxidative DNA damage⁴³ as also observed in our study in response to chronic hypoxic exposure. Thus, an additional mechanism by which Piezo1 deficiency protects against PH is the maintenance of distal pulmonary capillaries, linked to the reduction in the number of senescent ECs through inhibition of HPV and pulmonary vascular remodeling. The proposed model for the role of PASMC-Piezo1 in PH development is schematically summarized in Figure 8.

Figure 8.

Proposed mechanism for the role of pulmonary arterial smooth muscle cell (PASMC)-Piezo1 in the development and progression of pulmonary hypertension. EC indicates endothelial cell; and HPV, hypoxic pulmonary vasoconstriction.

By acting on these mechanisms, targeting Piezo1 in SMCs may, therefore, exert benefits in most forms of PH. The ability to affect Piezo1 in vivo by pharmacological means, however, remains limited at present: the low affinity, poor solubility, and stability of the available Piezo1 modulators still preclude in vivo administration. Moreover, the broad expression profile of Piezo1 in a variety of tissues and cell types, including ECs, greatly complicates possible systemic pharmacological treatment strategies. In conclusion, SMC-Piezo1 plays a key role in PH by influencing HPV and cell proliferation, but its specific pharmacological targeting remains a challenge.

ARTICLE INFORMATION

Acknowledgments

The authors thank N. Schupp and K. Malkmus for technical assistance and C. Veith for careful proofreading of the article. Biospecimen/data were provided by the UGMLC Giessen Biobank, member of the DZL Platform Biobanking. The authors are grateful to A. Lalot and D. Gelperowic from the Animal Facility and X. Decrouy, L. Wingertsmann, and W. Verbecq-Morlot from the Imaging Facility from the Institut Mondor de Recherche Biomédicale (IMRB).

Sources of Funding

This study was supported by grants from the French National Institute of Health and Medical Research, Agence Nationale de la Recherche (ANR)-20-CE14-0032-01 (Hypertension Artérielle Pulmonaire [HTAP]), the Fondation pour la recherche médicale, and the German Research Foundation, Cardio-Pulmonary Institute, EXC 2026, project ID 390649896.

Disclosures

None.

Supplemental Material

Supplemental Materials and Methods

Tables S1–S2

Figures S1–S7

Major Resources Table

References 44–47