Mechanosensitive cation currents through TRPC6 and Piezo1 channels in human pulmonary arterial endothelial cells

Abstract

Mechanosensitive cation channels and Ca²⁺ influx through these channels play an important role in the regulation of endothelial cell functions. Transient receptor potential canonical channel 6 (TRPC6) is a diacylglycerol-sensitive nonselective cation channel that forms receptor-operated Ca²⁺ channels in a variety of cell types. Piezo1 is a mechanosensitive cation channel activated by membrane stretch and shear stress in lung endothelial cells. In this study, we report that TRPC6 and Piezo1 channels both contribute to membrane stretch-mediated cation currents and Ca²⁺ influx or increase in cytosolic-free Ca²⁺ concentration ([Ca²⁺]_cyt) in human pulmonary arterial endothelial cells (PAECs). The membrane stretch-mediated cation currents and increase in [Ca²⁺]_cyt in human PAECs were significantly decreased by GsMTX4, a blocker of Piezo1 channels, and by BI-749327, a selective blocker of TRPC6 channels. Extracellular application of 1-oleoyl-2-acetyl-sn-glycerol (OAG), a membrane permeable analog of diacylglycerol, rapidly induced whole cell cation currents and increased [Ca²⁺]_cyt in human PAECs and human embryonic kidney (HEK)-cells transiently transfected with the human TRPC6 gene. Furthermore, membrane stretch with hypo-osmotic or hypotonic solution enhances the cation currents in TRPC6-transfected HEK cells. In HEK cells transfected with the Piezo1 gene, however, OAG had little effect on the cation currents, but membrane stretch significantly enhanced the cation currents. These data indicate that, while both TRPC6 and Piezo1 are involved in generating mechanosensitive cation currents and increases in [Ca²⁺]_cyt in human PAECs undergoing mechanical stimulation, only TRPC6 (but not Piezo1) is sensitive to the second messenger diacylglycerol. Selective blockers of these channels may help develop novel therapies for mechanotransduction-associated pulmonary vascular remodeling in patients with pulmonary arterial hypertension.

INTRODUCTION

Lung vascular endothelial cells (ECs) form the pulmonary arterial and venous endothelium (or the intima) and the capillaries around alveoli. Similar to ECs in systemic vasculature, lung ECs are considered nonexcitable cells but abundantly express cation and anion channels in the plasma membrane. In addition to regulating membrane potential, ion flux through channels in ECs may directly contribute to the regulation of production and release of vasoactive substances (e.g., endothelium-derived relaxing and constricting factors), mitogenic and angiogenic factors, and inflammatory cytokines. In addition, ion channels (e.g., Ca²⁺-permeable cation channels) in lung ECs are involved in trans capillary transportation of macromolecules through regulation of endocytosis, trans cytosine, and exocytosis (1, 2).

Ca²⁺-permeable cation channels in lung vascular ECs play an important role in the regulation of intracellular Ca²⁺ homeostasis, membrane potential, and various vascular function (1, 2). An increase in cytosolic-free Ca²⁺ concentration ([Ca²⁺]_cyt) in pulmonary arterial ECs (PAESs) is a trigger to activate endothelial nitric oxide (NO) synthase (eNOS) and stimulate endothelial NO production. By activating soluble guanylate cyclase to produce cGMP and/or activating various K⁺ channels to cause membrane hyperpolarization in smooth muscle cells (SMCs), NO is an important endothelium-derived relaxing factor (EDRF) that causes endothelium-dependent vasodilation (3–5). Persistent activation of eNOS due to EC calveolin-1 deficiency is, however, implicated in the development of pulmonary hypertension (PH) due to potentially eNOS uncoupling (6) and protein kinase G (PKG) nitration (7–10).

Vascular ECs including PAECs are nonexcitable cells that express both voltage-dependent Ca²⁺ channels (VDCC) (11, 12) and voltage-independent and/or receptor-operated cation channels (1). Changes of the activity, expression, and gating of these nonselective cation channels contribute to the regulation of EC function and vascular tone and structure (1, 13). Among the cation channels in ECs (including PAECs), the Ca²⁺-permeable nonselective cation channels [e.g., Transient receptor potential vanilloid 1 (TRPV1)] (14, 15) and Ca²⁺-activated K⁺ (K_Ca) channels (16, 17) are two important families of channels participating in the regulation of Ca²⁺ signaling and Ca²⁺-dependent translational and transcriptional cascades in vascular ECs. In addition to the rise in [Ca²⁺]_cyt due to Ca²⁺ influx through nonselective cation channels, Na⁺ influx through these channels plays an important role in regulating EC migration and mobility (18).

More than 40% of cells in the lungs are arterial, arteriole, capillary, and venous ECs (19), although the lung vasculature or vascular wall is composed of ECs, smooth muscle cells (SMCs), and (myo)fibroblasts. In whole lung tissues and isolated pulmonary artery (PA) rings, we and others observed upregulation of transient receptor potential (TRP) channels in animals with experimental PH (20–22). In pulmonary arterial smooth muscle cells (PASMC) and PAEC from patients with pulmonary arterial hypertension (PAH), we and others found that transient receptor potential canonical channel 6 (TRPC6) and Piezo1 are upregulated and pathogenically involved in the development of sustained pulmonary vasoconstriction and concentric pulmonary vascular remodeling (23–25). TRPC6 is a canonical TRP channel subunit (with a single-channel conductance of 35 pS) that participates in forming receptor-operated Ca²⁺ channels in vascular SMCs and other types of cells including ECs (26–29) that are activated by ligand-mediated activation of G protein-coupled receptors and tyrosine kinase receptors (30–33) and by redox changes (34). In addition, TRPC6 is also a mechanosensitive cation channel that allows Ca²⁺ and Na⁺ to enter cell (35–37). Piezo1 is a newly identified mechanosensitive cation channel that is involved in the regulation of blood pressure, vascular reactivity, and angiogenesis (38–43), as well as cell proliferation (44–47).

In this study, we examine whether TRPC6 and Piezo1 channels are involved in receptor-operated Ca²⁺ influx (or cation currents) and mechanical stimulation-induced Ca²⁺ influx (or cation currents) in human PAECs. We compared the cation currents in TRPC6-transfected and Piezo1-transfected Human embryonic kidney (HEK)-293 cells with the native human PAECs. The data from this study demonstrate that both TRPC6 and Piezo1 channels contribute to the whole cell cation currents induced by mechanical stimulation or membrane stretch. TRPC6 also contributes to the whole cell cation currents induced by the second messenger, diacylglycerol, or its membrane-permeable analog 1-oleoyl-2-acetyl-sn-glycerol (OAG), but Piezo1-associated mechanosensitive currents are not sensitive to OAG. These results suggest that, in addition to other cation channels expressed in lung ECs, TRPC6 is an important OAG-sensitive receptor-operated or diacylglycerol-activated cation channel, whereas both TRPC6 and Piezo1 are important mechanosensitive cation channels in human PAECs.

MATERIALS AND METHODS

Cell Preparation and Culture

Normal human PAECs (Lonza, Rockville, MD) were cultured in VascuLife VEGF growth medium (Lifeline Cell Technologies, Frederick, MD) supplemented with 2% fetal bovine serum (FBS), 5 ng/mL human epidermal growth factor (EGF), 5 ng/mL human fibroblast growth factor (FGF), 50 µg/mL ascorbic acid, 1 µg/mL hydrocortisone hemisuccinate, 10 mM l-glutamine, 15 ng/mL human insulin-like growth factor (IGF-1), 5 ng/mL human vascular endothelial growth factor (VEGF), 0.75 U/mL heparin sulfate, 30 mg/mL gentamicin, and 15 µg/mL amphotericin B. HEK-293 cells were cultured in high-glucose DMEM (Invitrogen) supplemented with 10% FBS (Invitrogen, Waltham, MA), 100 IU/mL penicillin and 100 μg/mL streptomycin (Sigma-Aldrich, St. Louis, MO). Cells were maintained in an incubator under a humidified environment at 37°C and 5% CO₂. Upon reaching 70%–90% confluence, cells were detached with Trypsin-EDTA (0.05%) and plated on 25-mm cover slips for electrophysiological and fluorescent microscopy experiments.

Transfection

HEK-293 cells were transiently transfected with 2 µg of human TRPC6 construct (Plasmid No. 21084; Addgene) or mouse Piezo1 construct (Plasmid No. 80925, Addgene) using X-tremeGENE HP DNA Transfection Reagent (Millipore Sigma, Burlington, MA) for 4–6 h in Opti-MEM Reduced Serum Medium (Gibco, Grand Island, NY). Then, the Opti-MEM medium was replaced with 10% FBS DMEM and the cells were incubated in 10% FBS DMEM for 48–72 h before being used for electrophysiological and digital imaging fluorescence microscopy experiments. A GFP construct was used to visualize the transfected cells by co-transfecting the cells with the TRPC6 construct or the Piezo1 construct. In the pCMS-EGFP vector, the EGFP gene was expressed separately from the gene of interest and is used as a transfection marker; most of the cells would be co-transfected efficiently with the GFP and the TRPC6 or Piezo1 gene. Green fluorescence emitted at 507 nm was used to visualize the cells transfected with TRPC6 or Piezo1 and GFP constructs using an inverted Nikon microscope (Eclipse/TE200) with the TE-FM epifluorescence attachment. The cell images were acquired with an Image Intensifier Tube/Philips 1381 system (Stanford Photonics Electronic Imaging Technologies, Palo Alto, CA). We also used siRNA to specifically downregulate TRPC6 in human PAECs. TRPC6-siRNA (Ambion s14422) was introduced into human PAECs using RNA Max Lipofectamine reagent (Invitrogen, Waltham, MA) with a final concentration of 100 pM per well in 6-well Petri dishes. The cells were incubated in the siRNA-transfection media for 48–72 h before being used for the patch clamp and digital imaging fluorescence microscopy experiments.

Electrophysiological Recording

Whole cell cation currents were recorded with the patch-clamp technique using an Axopatch 200B amplifier and a DigiData1550B interface (Molecular Devices, Silicon Walley, CA). Patch pipettes (2–3 MΩ) were fabricated on an electrode puller (Sutter Instrument, Novato, CA) using borosilicate glass tubes and fire polished on a microforge (Narishige Scientific Instruments, Tokyo, Japan) to get a range of 4–9 MΩ series resistance. Command voltage protocols and data acquisition were performed using pCLAMP-10 software (Molecular Devices). The HEPES-buffered bath (extracellular) solution contained 137 mM NaCl, 5.9 mM KCl, 1.8 mM CaCl₂, 1.2 mM MgCl₂, 14 mM glucose, and 10 mM HEPES (pH, 7.4, by NaOH). The pipette (intracellular) solution contained 120 mM CsOH, 120 mM aspartic acid, 0.6475 mM CaCl₂, 4 mM MgCl₂, 5 mM MgATP, 10 mM HEPES, and 10 mM EGTA (pCa = 8.0). The pH of the pipette solution was adjusted to 7.2 with 1 N CsOH. The hypotonic bath solution contained 90 mM NaCl, 5 mM KCl, 2.4 mM CaCl₂, 1.3 mM MgCl₂, 10 mM glucose, and 10 mM HEPES (pH = 7.4 with NaOH). The cell was clamped at a holding potential of 0 mV and an ascending ramp protocol from −100 to +100 mV for 1,000 ms was applied to the cell every 1.3 s. The step protocol has a 0-mV holding potential with 500-ms step in 200-mV increments from −100 to +100 mV. Currents were filtered at 1–2 kHz and digitized at 2–5 kHz. Electrophysiological recordings were performed at room temperature (24–25°C).

Measurement of Cytosolic Ca²⁺ Concentration ([Ca²⁺]_cyt)

Briefly, human PAECs were grown to 60%–70% confluence on 25-mm round glass coverslips. The cells were loaded with fura-2 acetoxymethyl ester (fura-2/AM, 4 µM; Invitrogen/Molecular Probes, Eugene, OR) in the dark for 60 min at room temperature (22–24°C) in a normal physiological salt solution (PSS). The PSS contained 140 mM NaCl, 4.7 mM KCl, 1.8 mM CaCl₂, 1.2 mM MgCl₂, 10.0 mM glucose, and 10.0 mM HEPES. A coverslip containing fura-2/AM-loaded cells was placed in a recording chamber on the stage of the Nikon inverted fluorescence microscope (Eclipse Ti-E; Nikon, Tokyo, Japan). The excitation wavelengths were 340 nm and 380 nm, and the emission signal at 520 nm was detected using an EM-CCD camera (Evolve; Photometrics, Tucson, AZ) through a Nikon S-Plan Fluor ELWD ×20/0.45 objective lens; the fluorescence images were recorded and analyzed using the NIS Elements 3.2 software (Nikon, Tokyo, Japan). [Ca²⁺]_cyt within the region of interest (5 × 5 μm) was positioned at the peripheral region of each cell and measured as the ratio of fluorescence intensities (F₃₄₀/F₃₈₀) and recorded every 2 s. [Ca²⁺]_cyt was calculated by the ratiometric method using the following equation: [Ca²⁺]_cyt = K_d × (S_f2/S_b2) × (R − R_min)/(R_max − R), where K_d (225 nM) is the dissociation constant of Fura-2 for Ca²⁺ and S_f2 and S_b2 are the fluorescent intensities at 380-nm excitation in Ca²⁺-free solution (with EGTA) and Ca²⁺-saturated solution (with ionomycin), respectively. R is the measured fluorescence ratio, while R_min and R_max are the minimum and maximum ratios that were determined in cells superfused with the Ca²⁺-free solution (plus 5 mM EGTA) and the Ca²⁺-containing solution (10 mM CaCl₂), respectively, in the presence of ionomycin (2 µM). In Ca²⁺-free solution, CaCl₂ was replaced by equimolar amounts of MgCl₂, and EGTA was added to chelate residual Ca²⁺. All experiments for measurement of [Ca²⁺]_cyt were performed at room temperature (22–24°C).

Drugs and Chemicals

All drugs and chemicals were obtained from Sigma Aldrich unless otherwise stated. A hydrophobic compound, 1-oleoyl-2-acetyl-sn-glycerol (OAG) which is a membrane-permeant analog of the second messenger, diacylglycerol (DAG), was dissolved in dimethyl sulfoxide (DMSO) at the concentration of 100 mM as a stock solution. BI-749327 (Cat. No. HY-111925, MedChemExpress, Monmouth, NJ) was prepared as an 11 mM stock solution in DMSO. GsMTx4 (Cat. No. HY-P1410, MedChemExpress, Monmouth, NJ) was dissolved into DMSO and prepared as a 500 mM stock solution. Yoda-1, 2-(5-{[(2,6-dichlorophenyl)methyl]sulfanyl}-1,3,4-thiadiazol-2-yl)pyrazine (Cat. No. 448947-81-7, Sigma-Aldrich, St. Louis, MO), was dissolved into DMSO and prepared as 30 mM stock solution. Aliquots of the stock solutions were then diluted to make a final concentration of OAG (Cat. No. 86390-77-4, Sigma-Aldrich, St. Louis, MO) at 100 µM, BI-749327 at 50 nM, and GsMTx4 at 5 µM in PSS on the day of the experiment.

Statistical Analysis

Pooled data are shown as means ± SE. The statistical significance between two groups was determined by Student’s t test. The statistical significance among groups was determined by one-way ANOVA analysis. A significant difference is expressed in the figures or figure legends by *P < 0.05, **P < 0.01, and ***P < 0.001.

RESULTS

We first conducted a series of patch clamp experiments to measure whole cell nonselective cation currents (I_Cation) in human PAECs induced by membrane stretch to demonstrate mechanosensitive cation currents. Lowering osmolality or decreasing osmotic pressure from 310 mosmol/kgH₂O (Iso-Osm, iso-osmalality or iso-osmotic solution) to 220 mosmol/kgH₂O (Hypo-Osm, hypo-osmolality or hypo-osmotic solution) in extracellular perfusate causes the movement of water molecules from the extracellular site into the cytosol and results in membrane stretch. The Hypo-Osm-mediated I_Cation is defined as the mechanosensitive cation current. OAG (a membrane-permeable analog of diacylglycerol) was used to induce diacylglycerol-sensitive cation currents in PAECs and the OAG-mediated I_Cation is defined as the receptor-operated cation current. Then, we used BI-749327, a selective blocker of TRPC6 (48), to characterize the pharmacological property of TRPC6 channels, and GsMTx4, a blocker of Piezo1 channels (49–51) and Yoda1, a selective activator of Piezo1 channels (52), to characterize Ca²⁺ influx through Piezo1 channels in human PAECs.

Membrane Stretch Induces Nonselective Currents in Human Pulmonary Artery Endothelial Cells

Membrane stretch by lowering extracellular osmolality from 310 mosmol/kgH₂O (Iso-Osm) to 220 mosmol/kgH₂O (Hypo-Osm) significantly and reversibly enhanced I_Cation in human PAECs (Fig. 1Aa). The membrane stretch-mediated I_Cation (Fig. 1Ab) was obtained by subtracting the current recorded when the cell was bathed in Iso-Osm solution from the current recorded when the cell was bathed in Hypo-Osm solution, which was reversed at ∼0 mV. The reversal potential of 0 mV indicated that these currents were nonselective cation currents. These data indicate that membrane stretch induces nonselective cation currents through, potentially, mechanosensitive cation channels in the PAEC plasma membrane. We then used a specific blocker of TRPC6 channels, BI-749327 (with an IC₅₀ of 19 nM for TRPC6, which is 40–80 times more potent than on other TRPC channels) (48), to examine whether the membrane stretch-mediated I_Cation were due to, at least partially, cation influx through TRPC6 channels in PAECs.

Figure 1.

Membrane stretch and diacylglycerol (DAG) induce nonselective cation currents (I_Cation) in human pulmonary arterial endothelial cells (PAECs) and BI-749327, a selective blocker of TRPC6, inhibits the stretch- and DAG-mediated cation currents. A: representative whole cell I_Cation (a), elicited by a ramp protocol from −100 mV to +100 mV (for 1,000 ms) in a human PAEC before (iso-osmalality or iso-osmotic solution, Iso-Osm, 330 mosmol/kgH₂O), during (hypo-osmolality or hypo-osmotic solution, Hypo-Osm), and after (Recovery) application of hypo-osmotic (Hypo-Osm; 220 mosmol/kgH₂O) solution. The stretch-mediated I_Cation (b) was constructed by subtracting I_Cation recorded in Iso-Osm from I_Cation recorded in Hypo-Osm. Summarized data (means ± SE, n = 5 cells, c) showing the I_Cation amplitudes at −100 mV and +100 mV in cells before, during, and after application of Hypo-Osm solution. **P < 0.01, ***P < 0.001 vs. Iso-Osm and Recovery (one-way ANOVA). B: representative I_Cation (a), elicited by a ramp protocol in a PAEC bathed in Hypo-Osm solution before (Control), during (BI-749327), and after (Washout) application of BI-749327 (50 nM) in Hypo-Osm solution. The BI-749327-sensitive stretch-mediated I_Cation (b) was constructed by subtracting I_Cation recorded in Control + Hypo-Osm from I_Cation recorded in BI-749327 + Hypo-Osm. Summarized data (means ± SE, n = 5 cells, c) showing current amplitudes at −100 mV and +100 mV in PAECs in Hypo-Osm before, during, and after application of BI-749327. **P < 0.01, ***P < 0.001 vs. Control and Washout (one-way ANOVA). C: representative I_Cation (a) in a PAEC before (Control), during (1-oleoyl-2-acetyl-sn-glycerol, OAG), and after (Washout) application of the DAG analog OAG (100 µM). The OAG-mediated I_Cation (b) was constructed by subtracting the current recorded before application of OAG from the current recorded during application of OAG. Summarized data (means ± SE, n = 5 cells, c) showing the amplitudes of I_Cation at −100 mV and +100 mV in PAECs before, during, and after extracellular application of OAG (100 µM). **P < 0.01, ***P < 0.001 vs. Control and Washout (one-way ANOVA). D: representative I_Cation (a) in a PAEC bathed in OAG-containing Iso-Osm solution before, during, and after application of BI-749327 (50 nM). The BI-749327-sensitive OAG-mediated I_Cation (b) was constructed by subtracting I_Cation recorded in OAG from I_Cation recorded in OAG + BI-749327. Summarized data (means ± SE, n = 5 cells, c) showing the amplitudes of currents at −100 mV and +100 mV in PAECs in OAG solution before, during, and after application of BI-749327 (50 nM). ***P < 0.001 vs. Control (one-way ANOVA).

In the cells bathed in Hypo-Osm solution, the plasma membrane was stretched by hypo-osmolality (as a result of inward water transportation) and the mechanosensitive I_Cation were induced by the membrane stretch. Extracellular application of BI-749327 significantly and reversibly inhibited the stretch-mediated increases in whole cell I_Cation (Fig. 1Ba). The BI-749327-sensitive membrane stretch-mediated I_Cation (Fig. 1Bb) obtained by subtracting the stretch-induced current recorded in the absence of BI-749327 (Control) from the stretch-induced current recorded in the presence of 50 nM BI-749327 also reversed at 0 mV with a little inward rectification at the high voltage (+60 to +100 mV (Fig. 1Bb). BI-749327 resulted in an ∼60% inhibition of the stretch-induced I_Cation at −100 mV and +100 mV (Fig. 1Bc). These results indicate that, in human PAECs, membrane stretch activates nonselective cation channels, whereas BI-749327-sensitive TRPC6 channels contribute to the mechanosensitive I_Cation.

Diacylglycerol Analog Induces Nonselective Currents in Human PAECs

In addition to mechanosensitive stimulation (i.e., membrane stretch by extracellular hypo-osmolality), extracellular application of OAG (100 µM), a membrane-permeable analog of the second messenger diacylglycerol (DAG), significantly and reversibly enhanced the whole cell I_Cation in human PAECs (Fig. 1Ca). The reversal potential and current kinetics of the OAG-sensitive I_Cation (Fig. 1Cb), obtained by subtracting the current recorded before OAG treatment (Control) from the current recorded during OAG treatment, are similar to the stretch-mediated I_Cation (shown in Fig. 1Ab). In the presence of 100 µM of OAG, extracellular application of BI-749327 (50 nM), a selective blocker of TRPC6 channels, significantly and reversibly reduced the OAG-mediated I_Cation (Fig. 1Ca). The BI-749327-sensitive OAG-mediated I_Cation (Fig. 1Db), obtained by subtracting the OAG-induced I_Cation recorded in the absence of BI-749327 (Control) from the OAG-induced I_Cation recorded in the presence of 50 nM BI-749327, seems to have similar reversal potential (0 mV) and current kinetics to the BI-749327-sensitive stretch-mediated I_Cation (Fig. 1Bb). These data indicate that, in human PAECs, the second messenger DAG or the exogenous analog OAG activates nonselective cation channels, whereas BI-749327-sensitive TRPC6 channels contribute to the OAG-induced I_Cation.

Downregulation of TRPC6 Inhibits Stretch-Induced Cation Currents in Human PAECs

To further confirm the contribution of TRPC6 to the mechanosensitive I_Cation, we treated human PAECs with siRNA specifically targeting TRPC6 (TRPC6-siRNA) and compared the stretch-mediated I_Cation with control cells. As shown in Fig. 2, Hypo-Osm solution resulted in a significant increase in whole cell I_Cation in control PAECs (Fig. 2, A and B, left). The averaged amplitude of mechanosensitive I_Cation in control PAECs was −144 pA at −100 mV and 214 pA at +100 mV (Fig. 2, B and C). Treatment of PAECs with TRPC6-siRNA (for 48 h) significantly reduced the Hypo-Osm-mediated I_Cation (Fig. 2, A and B, right); the averaged amplitude of mechanosensitive I_Cation in TRPC6-siRNA-treated PAECs was −60.8 pA at −100 mV and 33.4 pA at +100 mV (Fig. 2, B and C, left). The approximate 80% of inhibition of stretch-mediated I_Cation at −100 and +100 mV (Fig. 2C, right) indicate that cation currents through TRPC6 channels account for a large portion of mechanosensitive cation currents in human PAECs.

Figure 2.

Downregulation of transient receptor potential canonical channel 6 (TRPC6) with siRNA significantly inhibits membrane stretch-induced cation currents (I_Cation) in human pulmonary arterial endothelial cell (PAECs). A: representative whole cell I_Cation, elicited by a ramp protocol from −100 mV to + 100 mV (for 1,000 ms) before (iso-osmalality or iso-osmotic solution, Iso-Osm), during (hypo-osmolality or hypo-osmotic solution, Hypo-Osm), and after (Recovery) application of Hypo-Osm solution in a control PAEC (Control, left) and a PAEC treated with TRPC6-siRNA (100 pM/well for 48 h, right). B: summarized data (means ± SE, n = 5 cells showing I_Cation amplitudes at −100 mV and +100 mV before, during, and after application of Hypo-Osm solution in control PAECs (left) and TRPC6-siRNA-treated PAECs (right). ***P < 0.001 vs. Iso-Osm and Recovery (one-way ANOVA). C: the stretch-mediated I_Cation shown in the left panel was constructed by subtracting I_Cation recorded in Iso-Osm from I_Cation recorded in Hypo-Osm in control PAEC and TRPC6-siRNA-treated PAEC (A). Summarized data (means ± SE, n = 5, right) showing the percentage change of the amplitude of I_Cation (at −100 mV and +100 mV) in TRPC6-siRNA-treated PAEC compared to control cells.

TRPC6 Channels Are Sensitive to Membrane Stretch Due to Hypo-osmolality

The mechanosensitivity of TRPC6 has been demonstrated in different types of cells (35, 53–55). To further prove that TRPC6 channels are sensitive to membrane stretch, we conducted similar experiments in HEK-cells transiently transfected with the human TRPC6 gene. As shown in Fig. 3, membrane stretch, by superfusing hypo-osmotic solution (Hypo-Osm, for 10 min), significantly and reversibly increased whole cell nonselective cation currents through TRPC6 channels (I_TRPC6) in TRPC6-transfected HEK cells (Fig. 3A, left). The membrane stretch-mediated I_TRPC6 (Fig. 1A, right), obtained by subtracting the current recorded when the cell was bathed in Iso-Osm solution from the current recorded when the cell was bathed in Hypo-Osm solution, reversed at ∼0 mV. In the cells in which the currents were elicited by step-pulse, from a holding potential of 0 mV to +100 mV or to −100 mV, membrane stretch also significantly and reversibly enhanced the outward (at+100 mV) and inward (at −100 mV) currents (Fig. 3, B and C). These results indicate that membrane stretch of PAEC due to hypo-osmolality is sufficient to activate TRPC6 channels and that the stretch-mediated cation currents through TRPC6 are reversible upon restoration of extracellular osmolality.

Figure 3.

Membrane stretch and diacylglycerol (DAG) induce cation currents through transient receptor potential canonical channel 6 (TRPC6) channels (I_TRPC6) in TRPC6-transfected human embryonic kidney (HEK)-293 cells and BI-749327 inhibits the stretch-and diacylglycerol-mediated I_TRPC6. A: representative whole cell I_TRPC6 (left), elicited by a ramp protocol from −100 mV to + 100 mV (for 1,000 ms) in a TRPC6-transfected HEK cell before (iso-osmalality or iso-osmotic solution, Iso-Osm), during (Hypo-osmolality or hypo-osmotic solution, Hypo-Osm), and after (Recovery) application of Hypo-Osm solution. The stretch-mediated I_TRPC6 (right) was constructed by subtracting the current recorded in Iso-Osm from the current recorded in hypo-osmolality or hypo-osmotic solution (Hypo-Osm). B: representative whole cell I_TRPC6, elicited by depolarization from a holding potential of 0 mV to +100 mV and −100 mV in a TRPC6-transfected HEK cell before, during, and after application of Hypo-Osm solution. C: summarized data (means ± SE, n = 5 cells) showing the amplitudes of I_TRPC6 at −100 mV and +100 mV in TRPC6-transfected HEK cells before, during, and after application of Hypo-Osm solution. *P < 0.05, ***P < 0.001 vs. Iso-Osm and Iso-Osm Recovery (one-way ANOVA). D: representative I_TRPC6 (left), elicited by a ramp protocol in a TRPC6-transfected HEK cell before (Control), during (1-oleoyl-2-acetyl-sn-glycerol, OAG) and after (Washout) application of the DAG analog OAG (100 µM). The OAG-mediated I_TRPC6 (right) was constructed by subtracting the current recorded in Control and from the current recorded in OAG. E: representative I_TRPC6, elicited by depolarization from a holding potential of 0 mV to +100 mV and −100 mV in a TRPC6-transfected HEK cell before, during, and after application of OAG (100 µM). F: summarized data (means ± SE, n = 5 cells) showing the amplitudes of I_TRPC6 at −100 mV and +100 mV in TRPC6-transfected HEK cells before, during, and after application of OAG. ***P < 0.001 vs. Control and Washout (one-way ANOVA).

TRPC6 Channels Are Activated by the Diacylglycerol Analog

In addition to its mechanical sensitivity (53), TRPC6 is known as an important receptor-operated Ca²⁺ channel that is activated by ligands and the second messenger diacylglycerol (DAG) (26, 30, 36). Similar to human PAECs, extracellular application of the membrane-permeable DAG analog, OAG (100 µM), significantly and reversibly increased whole cell cation currents in TRPC6-transfected HEK cells (Fig. 3D, left). The OAG-sensitive currents, obtained by subtracting the currents recorded in the cell superfused with OAG-containing solution from the currents recorded in the cell bathed in control solution (Fig. 3D, right) indicate that the cation currents through TRPC6 channels (or predominantly through TRPC6 channels) seem to have a similar current-voltage (I-V) relationship as the native cation currents activated by OAG in human PAECs (Fig. 1C, a and b). In addition to the ramp protocol, we also used step protocol to elicit the currents by depolarizing the cell from a holding potential of 0 mV to +100 mV (to induce outward current) and −100 mV (to induce inward current). Extracellular application of 100 µM OAG resulted in a 162% increase of the inward current at −100 mV and a 97% increase of the outward currents at +100 mV in TRPC6-transfected HEK cells (Fig. 3, E and F). These results are in good agreement with other reports that the membrane-permeable DAG analog, OAG, can directly activate TRPC6 channels.

BI-749327 Significantly Reduces Cation Currents through TRPC6 Channels

In control HEK cells (Fig. 4A) and HEK cells treated with transfection agents (Mock, Fig. 4B), extracellular application of 50 nM BI-749237 had negligible effect on the “basal” cation currents elicited by either a ramp protocol (a and b) or a step protocol (c and d). Transient transfection of the human TRPC6 gene significantly enhanced the cation currents (Fig. 4C) compared with control HEK cells (Fig. 4A) and Mock HEK cells (Fig. 4B). In TRPC6-transfected HEK cells, extracellular application of BI-749327 (50 nM) significantly and reversibly inhibited the cation currents through TRPC6 channels (I_TRPC6) (Fig. 4C) without membrane stretch or stimulation with OAG. The BI-749327-sensitive whole cell I_TRPC6, obtained by subtracting the currents recorded in the cell superfused with solution containing 50 nM BI-749327 from the currents recorded in the cell bathed in control (C) solution (Fig. 5Cb) shows that the currents reversed at ∼0 mV and the I-V relationship is similar to the TRPC6 currents demonstrated by other investigators (56). Using both the ramp (Fig. 4C, a and b) and step (Fig. 4C, c and d) protocols, we showed that 50 nM BI-749327 resulted in a 62%–71% decrease of the inward I_TRPC6 at −100 mV and an 81%–85% decrease of the outward I_TRPC6 at +100 mV (Fig. 5C, c and d). The next set of experiments was designed to examine whether BI-749327-mediated inhibition of TRPC6 channels inhibits OAG-mediated increases in [Ca²⁺]_cyt in TRPC6-transfected HEK cells. Consistent with the patch clamp experiments, extracellular application of 50 nM BI-749327 negligibly affected the resting [Ca²⁺]_cyt, but significantly inhibited OAG-induced increases in [Ca²⁺]_cyt (Fig. 5, A and B). These data indicate that BI-749327 is a potent inhibitor of TRPC6 channels, and BI-749327-mediated blockade of TRPC6 causes significant inhibition of receptor-operated or OAG-mediated cation currents and Ca²⁺ entry.

Figure 4.

BI-749327 inhibits cation currents through transient receptor potential canonical channel 6 (TRPC6) channels (I_TRPC6) in TRPC6-transfected human embryonic kidney (HEK) cells. A and B: representative whole cell I_Cation (a), elicited by a ramp protocol from −100 mv to +100 mV (for 1,000 ms) in a control HEK cell (A) and a mock HEK cell (B, transfected with empty vector) before (C), during (BI), and after (W) application of BI-749327 (50 nM). The BI-749327-sensitive I_Cation (b) was constructed by subtracting the current in control (C) from the current recorded in BI-749327 (BI). Representative whole cell I_Cation (c), elicited by a step protocol from a holding potential of 0 mV to +100 mV and −100 mV in a control HEK cell (A) and a mock HEK cell (B) before (Control), during (BI-749327), and after (Washout) application of BI-749327. Summarized data (means ± SE, n = 5 cells, d) showing the amplitudes of I_Cation at −100 mV and +100 mV in control (A) and mock (B) HEK cells before, during, and after application of BI-749327. *P < 0.05 vs. Control (One-way ANOVA). C: representative whole cell I_TRPC6 (a), elicited by a ramp protocol in TRPC6-transfected HEK cell before (C), during (BI), and after (W) application of BI-749327. The BI-749327-sensitive I_TRPC6 (b) was constructed by subtracting I_TRPC6 recorded before application of BI-749327 (C) from I_TRPC6 recorded during application of BI-749327 (BI). Representative whole cell I_TRPC6 (c), elicited by a step protocol from a holding potential of 0 mV to +100 mV and −100 mV in a TRPC6-transfected HEK cell before, during, and after application of BI-749327. Summarized data (means ± SE, n = 5 cells, d) showing the amplitude of I_TRPC6 at −100 mV and +100 mV in TRPC6-transfected HEK cells before, during, and after application of BI-749327. ***P < 0.001 vs. Control (one-way ANOVA). I_cation, nonselective cation currents.

Figure 5.

BI-749327 (BI) reversibly inhibits 1-oleoyl-2-acetyl-sn-glycerol (OAG)-mediated increases in [Ca²⁺]_cyt in human pulmonary arterial endothelial cell (PAECs). A: representative record of changes in [Ca²⁺]_cyt before and during application of OAG (100 µM) in single PAECs (individual, a) and in group of PAECs (averaged, b) on the same cover slip bathed in physiological salt solution (PSS) (Control), PSS with BI-749327 (50 nM), and PSS with vehicle (Vehicle). B: summarized data (means ± SE, n = 29–32 cells) showing the basal [Ca²⁺]_cyt (left) and the OAG-induced increases in cytosolic-free Ca²⁺ concentration ([Ca²⁺]_cyt) (right), calculated from individual PAECs, in cells bathed in control PSS, PSS with BI-749327, and PSS with vehicle (DMSO). ***P < 0.001 vs. Control and Vehicle (one-way ANOVA).

The results from these experiments shown in Figs. 1, 2, 3, 4, and 5 indicate that the BI-749327-inhibited and OAG-activated TRPC6 channels are an important receptor-operated or diacylglycerol-activated cation channel in human PAECs that can be activated by mechanical stimulation. The sensitivity of TRPC6 channels in lung ECs to membrane stretch and second messenger make the channel as an important Ca²⁺ influx pathway for increasing [Ca²⁺]_cyt. Under normal conditions, the agonist- and stretch-mediated Ca²⁺ influx through TRPC6 channels in PAECs exerts vasodilative effect on the pulmonary vasculature by activating eNOS and increasing endothelial NO release. Under pathological conditions, the upregulated TRPC6 and enhanced Ca²⁺ influx through TRPC6 channels may be an important pathogenic mechanism for stretch- and mitogen-mediated pulmonary vascular remodeling.

The next set of experiments was designed to investigate whether Piezo1 channels are involved in generating mechanosensitive cation currents and increases in [Ca²⁺]_cyt in human PAECs.

GsMTx4 Inhibits Stretch-Mediated Nonselective Cation Currents in Human PAECs

In the cells bathed in Hypo-Osm solution or when membrane was stretched by hypo-osmolality, extracellular application of GsMTx4 significantly and reversibly inhibited the stretch-mediated increases in whole cell I_Cation (Fig. 6A, left). The GsMTx4-sensitive membrane stretch-mediated I_Cation (Fig. 6A, right), obtained by subtracting the stretch-induced current recorded in the absence of GsMTx4 (Control) from the stretch-induced current recorded in the presence of 5 µM GsMTx4, reversed at −20 mV (Fig. 6A, right). GsMTx4 resulted in an 40%–60% inhibition of the stretch-induced I_Cation at −100 mV and +100 mV (Fig. 6B). These results indicate that, in human PAECs, membrane stretch activates nonselective cation channels, whereas GsMTx4-sensitive Piezo1 channels also contribute to the mechanosensitive I_Cation.

Figure 6.

Membrane stretch induces cation currents (I_Cation) in human pulmonary arterial endothelial cell (PAECs) and GsMTx4 (a blocker of Piezo1) inhibits the stretch-mediated (I_Cation). A: representative whole cell I_Cation (left), elicited by a ramp protocol in a PAEC bathed in hypo-osmolality or hypo-osmotic solution (Hypo-Osm) solution before (Control), during (GsMTx4), and after (Washout) application of GsMTx4 (5 µM). The GsMTx4-sensitive stretch-mediated I_Cation (right) was constructed by subtracting I_Cation recorded in Control + Hypo-Osm from I_Cation recorded in GsMTx4 + Hypo-Osm. B: summarized data (means ± SE, n = 5 cells) showing the amplitudes of currents at −100 mV and +100 mV in PAECs bathed in Hypo-Osm before, during, and after application of GsMTx4. ***P < 0.001 vs. Control (one-way ANOVA).

We also compared the effects of GsMTx3 (a Piezo1 blocker) and BI-749327 (a TRPC6 blocker) on the increases in [Ca²⁺]_cyt induced by membrane stretch (Hypo-Osm) or Yoda-1, an allosteric activator of Piezo1 in human PAECs. Membrane stretch with Hypo-Osm solution induced a transient increase in [Ca²⁺]_cyt in PAECs (Fig. 7, A and B), the Hypo-Osm-mediated increase in [Ca²⁺]_cyt was significantly inhibited by the Piezo1 blocker GsMTx4 (GsM) (49) (Fig. 7C) and by the selective TRPC6 blocker BI-749327 (Fig. 7C). Extracellular application of Yoda-1 induced a transient increase in [Ca²⁺]_cyt followed by a sustained or plateaued increase in PAECs (Fig. 7, D and E). GsMTx4 significantly inhibited the Yoda-1-mediated increase in [Ca²⁺]_cyt, whereas BI-749327 had no effect on Yoda-1-mediated increase in [Ca²⁺]_cyt (Fig. 7F). These data imply that both GsMTx4-sensitive Piezo1 and BI-749327-sensitive TRPC6 contribute to the stretch-mediated increase in [Ca²⁺]_cyt in human PAECs, whereas only GsMTx4-sensitive Piezo1 contributes to Yoda-1-mediated increase in [Ca²⁺]_cyt.

Figure 7.

Effects of GsMTx4 (a Piezo1 blocker) and BI-749327 (a TRPC6 blocker) on membrane stretch- and Yoda-1-mediated increases in cytosolic-free Ca²⁺ concentration ([Ca²⁺]_cyt) in human Pulmonary arterial endothelial cell (PAECs). A: representative record showing [Ca²⁺]_cyt before and after membrane stretch induced by Hypo-Osm solution in the absence (Control, with vehicle) and presence of GxMTx4 (5 µM) or BI-749327 (50 nM). B: summarized data (means ± SE, n = 5) showing the basal level (Basal) of [Ca²⁺]_cyt and the peak and plateau levels of [Ca²⁺]_cyt during membrane stretch with Hypo-Osm solution in the absence and presence of GxMTx4 or BI-749327. ***P < 0.001 vs. Basal (one-way ANOVA). C: summarized data (means ± SE, n = 5) showing the transient (peak) increases of [Ca²⁺]_cyt induced by membrane stretch in the absence (Cont) and presence of GxMTx4 (GsM) or BI-749327 (BI). ***P < 0.001 vs. Cont (one-way ANOVA). D: representative record showing [Ca²⁺]_cyt before and after application of Yoda-1 (5 µM) in the absence and presence of GxMTx4 or BI-749327. E: summarized data (means ± SE, n = 5) showing the basal level (Basal) of [Ca²⁺]_cyt and the peak and plateau levels of [Ca²⁺]_cyt during application of Yoda-1 in the absence and presence of GxMTx4 or BI-749327. *P < 0.05, **P < 0.01, ***P < 0.001 vs. Basal (one-way ANOVA). F: summarized data (means ± SE, n = 5) showing the transient (peak) increases of [Ca²⁺]_cyt induced by Yoda-1 in the absence and presence of GxMTx4 or BI-749327. ***P < 0.001 vs. Cont (one-way ANOVA).

GsMTx4 Reduces Stretch-Induced Cation Currents but Not OAG-Induced Cation Currents in Piezo1-Transfected HEK Cells

In HEK cells treated with transfection agents (Mock, Fig. 8A), Hypo-Osm solution induced a small increase of cation currents elicited by either a ramp protocol (a and b). Transient transfection of the Piezo1 gene significantly enhanced not only the basal cation currents in HEK cells bathed in Iso-Osm solution (without membrane stretch or stimulation with OAG), but also the stretch-mediated cation currents in HEK cells bathed in Hypo-Osm solution (Fig. 8B, left). In Piezo1-transfected HEK cells bathed in Hypo-Osm solution, extracellular application of 5 µM GsMTx4 significantly inhibited the cation currents through Piezo1 channels (I_Piezo1) (Fig. 8B, right). The GsMTx4-sensitive whole cell I_Piezo1, obtained by subtracting the currents recorded in the cell superfused with solution containing 5 µM GsMTx4 from the currents recorded in the cell bathed in control (Control) Hypo-Osm solution (Fig. 8Bb, left) shows that the reversal potential of the currents is close to 0 mV and that the I-V relationship shows more outward rectification. GsMTx4 (5 µM) resulted in 55% decrease of the inward I_Piezo1 at −100 mV and 64% reduction of the outward I_Piezo1 at +100 mV (Fig. 8Bc, right) in Piezo1-transfected HEK cells. However, extracellular application of OAG had negligible effect on the whole cell I_Piezo1 in Piezo1-transfected HEK cells (Fig. 8C). The stretch-mediated enhancement of I_Piezo1 in Piezo1-transfected HEK cells are consistent with the Hypo-Osm-mediated increase in [Ca²⁺]_cyt. Membrane stretch by hypo-osmolality (or Hypo-Osm solution) induced a transient increase in [Ca²⁺]_cyt followed by a sustained increase in Piezo1-transfected HEK cells (Fig. 9A). By directly activating Piezo1 channels, extracellular application of Yoda-1 (5 µM) also resulted in a transient increase in [Ca²⁺]_cyt in Piezo1-transfected HEK cells (Fig. 9B). The pattern of Yoda1-mediated increase in [Ca²⁺]_cyt via Ca²⁺ influx through Piezo1 channels is similar to those reported by other investigators (52).

Figure 8.

GsMTX4 inhibits stretch-mediated cation currents through Piezo1 channels (I_Piezo1) in Piezo1-transfected human embryonic kidney (HEK) cells. A: representative whole cell I_Cation (a), elicited by a ramp protocol in a mock HEK cell (transfected with empty vector) before (iso-osmalality or iso-osmotic solution, Iso-Osm), during (hypo-osmolality or hypo-osmotic solution, Hypo-Osm), and after (Recovery) application of Hypo-Osm solution. The stretch-mediated I_Cation (b) was constructed by subtracting I_Cation recorded in Iso-Osm from I_Cation recorded in Hypo-Osm. Summarized data (means ± SE, n = 5 cells, c) showing the amplitude of I_Cation at −100 mV and +100 mV in mock HEK cells before, during, and after application of Hypo-Osm solution. ***P < 0.001 vs. Iso-Osm (one-way ANOVA). B: representative whole cell I_Piezo1 (a, left), elicited by a ramp protocol in a Piezo1-transfected HEK cell before, during, and after application of Hypo-Osm solution. The stretch-mediated I_Piezo1 (b, left) was constructed by subtracting I_Piezo1 recorded in Iso-Osm from I_Piezo1 recorded in Hypo-Osm. Summarized data (means ± SE, n = 5 cells, c, left) showing the amplitude of I_Piezo1 at −100 mV and +100 mV in mock HEK cells before, during, and after application of Hypo-Osm solution. ***P < 0.001 vs. Iso-Osm (one-way ANOVA). B, right: representative I_Piezo1 (a, right) in a Piezo1-transfected HEK cell bathed in Hypo-Osm solution before (Control), during (GsMTx4), and after (Washout) application of GsMTx4 (5 µM) in Hypo-Osm solution. The GsMTx4-sensitive stretch-mediated I_Piezo1 (b, right) was constructed by subtracting I_Piezo1 recorded in Control from I_Piezo1 recorded in GsMTx4. Summarized data (means ± SE, n = 5 cells, c, right) showing the amplitudes of currents at −100 mV and +100 mV in Piezo1-transfected HEK cells bathed in Hypo-Osm solution before, during, and after application of GsMTx4. ***P < 0.001 vs. Control (one-way ANOVA). C: representative I_Piezo1 (a) in a Piezo1-transfected HEK cell before (Control), during (1-oleoyl-2-acetyl-sn-glycerol, OAG), and after (Washout) application of the diacylglycerol analog OAG (100 µM). The OAG-mediated I_Piezo1 (b) was constructed by subtracting the current recorded in Control from the current recorded in OAG. Summarized data (means ± SE, n = 5 cells, c) showing the amplitudes of I_Piezo1 at −100 mV and +100 mV in Piezo1-transfected HEK cells before, during, and after application of OAG. I_cation, nonselective cation currents.

Figure 9.

Membrane stretch and Yoda-1 induce Ca²⁺ influx by activating Piezo1 channels. A: representative record (left) of changes in cytosolic-free Ca²⁺ concentration ([Ca²⁺]_cyt) in a Piezo1-transfected human embryonic kidney (HEK) cell before and during application of hypo-osmolality (Hypo-Osm) solution. Summarized data (means ± SE, n = 5 individual experiments) showing the basal [Ca²⁺]_cyt before Hypo-Osm and the peak and plateau increases in [Ca²⁺]_cyt during Hypo-Osm solution (right). ***P < 0.001 vs. Basal (one-way ANOVA). B: representative record (left) of changes in [Ca²⁺]_cyt in a Piezo1-transfected HEK cell before and during application of Yoda-1 (5 µM). Summarized data (means ± SE, n = 5 individual experiments) showing the basal [Ca²⁺]_cyt before Yoda-1 application and the peak and plateau increases in [Ca²⁺]_cyt during Yoda-1 application (right). **P < 0.01 vs. Basal (one-way ANOVA).

These data indicate that GsMTx4 is an inhibitor of Piezo1 channels and GsMTx-mediated blockade of Piezo1 causes significant inhibition of stretch-mediated cation currents through Piezo1 channels, whereas OAG had negligible effect on the cation currents through Piezo1 channels. Taken together, the data indicate that TRPC6 in human PAECs contributes to cation currents that are sensitive to the second messenger diacylglycerol (or OAG) and membrane stretch, whereas Piezo1 is a mechanosensitive cation channel in PAECs that is not directly activated by diacylglycerol (or OAG).

DISCUSSION

In this study, we characterized whole cell nonselective cation currents (I_Cation) in human PAECs that are induced or enhanced by membrane stretch and diacylglycerol. The observations showed that the currents generated by Ca²⁺ or cation flux through TRPC6 and Piezo1 channels contributed to the mechanosensitive cation currents in human PAECs (Fig. 10). Downregulation of TRPC6 with siRNA or blockade of TRPC6 with BI-749327, a potent and selective blocker of TRPC6 (48), significantly inhibited the stretch-mediated cation currents and the stretch-mediated increase in [Ca²⁺]_cyt in human PAECs. In HEK-293 cells transiently transfected with TRPC6 gene, OAG (a diacylglycerol analog) and membrane stretch (via superfusion of hypo-osmotic solution) both enhanced the cation currents through TRPC6 channels (I_TRPC6). Inhibition of Piezo1 channels with GsMTx4, a blocker of Piezo1 channels (49–51), also significantly reduced the stretch-mediated I_Cation and the stretch-mediated increase in [Ca²⁺]_cyt in human PAECs. In HEK-293 cells transiently transfected with the Piezo1 gene, membrane stretch enhanced the cation currents through Piezo1 channels (I_Piezo1) and GsMTx4 significantly reduced the stretch-mediated I_Piezo1, whereas OAG had negligible effect on I_Piezo1. The data from this study indicate that TRPC6 and Piezo1 are major cation channels involved in generating mechanosensitive cation currents in human PAECs. TRPC6 is not only a diacylglycerol-sensitive cation channel, but also a mechanosensitive cation channel in human PAECs. Piezo1, on the other hand, is a mechanosensitive channel that is insensitive to diacylglycerol (Fig. 10).

Figure 10.

Simplified schematic diagram illustrating the potential role of mechanosensitive cation channels of transient receptor potential canonical channel 6 (TRPC6) and Piezo1 in human pulmonary arterial endothelial cell (PAEC). Membrane stretch and/or flow shear stress activate both TRPC6 and Piezo1, as mechanosensitive cation channels, on the surface membrane of PAEC, enhance Ca²⁺ influx and raise cytosolic-free Ca²⁺ concentration ([Ca²⁺]_cyt). The stretch-mediated increase in [Ca²⁺]_cyt activates endothelial nitric oxide (NO) synthase (eNOS), increases NO synthesis and release, and causes endothelium-dependent pulmonary vasodilation. Increased [Ca²⁺]_cyt can also activate CaM/CaMK and AKT/mTOR signaling cascades to stimulate cell proliferation and gene expression, contributing to pulmonary vascular remodeling. Vasoconstrictor or mitogenic factor (e.g., endothelin-1, TXA₂) also opens TRPC6, as a receptor-operated cation channel, through diacylglycerol (DAG), to induce Ca²⁺ influx and contribute to regulating endothelium-dependent vasodilation and vascular remodeling. mTOR, mammalian target of rapamycin.

In the lungs, more than 40% of cells are endothelial cells (19) and the pulmonary vasculature or the pulmonary vascular endothelium is under constant stretch due to respiration and cardiac output. Under normal conditions, the stretch-mediated Ca²⁺ influx through mechanosensitive channels and agonist-induced Ca²⁺ influx through diacylglycerol-activated Ca²⁺ channels serve as an important mechanism for inducing the endothelium-derived relaxation by activating eNOS activity and NO production. In addition to its vasodilatative effect through activation of eNOS, the enhanced Ca²⁺ influx through mechanosensitive and receptor-operated cation channels (e.g., TRPC6 and Piezo1) in PAECs may contribute to the development and progression of PH by causing eNOS uncoupling (7, 57), activating the AKT/mammalian target of rapamycin (mTOR) signaling (25), and enhancing endothelial-to-mesenchymal transition (EndMT) (58, 59).

Mechanosensitive Ca²⁺ influx through Piezo1 channels in the endothelium is implicated in pulmonary vasodilation using isolated intrapulmonary arterial vessels (42). In the isolated and perfused/ventilated whole lung preparation, however, activation of Piezo1 channels with Yoda-1 caused an increase in pulmonary arterial pressure (PAP), whereas inhibition of eNOS with nitro-l-arginine methyl ester (l-NAME) significantly enhanced Yoda-1-mediated increase in PAP (data not shown). In addition, agonist (such as acetylcholine/ACh), by inducing Ca²⁺ influx through receptor-operated cation channels in the endothelium or ECs, causes pulmonary vasodilation, whereas functional removal (or injury) of endothelium from isolated pulmonary arterial rings converts the ACh-mediated pulmonary vasodilation to ACh-mediated vasoconstriction (60, 61) and significantly enhances vasoconstriction induced by other agonists (60, 62, 63). These data indicate that mechanosensitive and receptor-operated cation channels in ECs or PAECs are responsible for Ca²⁺ influx that activates eNOS and causes vasodilation. However, upregulation of the cation channels and persistent increases in [Ca²⁺]_cyt in ECs (due to Ca²⁺ influx through the upregulated mechanosensitive and receptor-operated cation channels) would activate other Ca²⁺- or Ca²⁺/CaM-sensitive signaling cascades (e.g., AKT/mTOR) to induce EndMT and EC proliferation, and ultimately contribute to the development of pulmonary vascular remodeling in patients with PAH and animals with experimental PH (25). The data from this study suggest that TRPC6 and Piezo1 channels are two important cation channels involved in raising [Ca²⁺]_cyt in human PAECs in response to mechanical stretch and flow shear stress reinforced by respiration and cardiac output.

Lung vascular injury associated with inflammation and oxidative stress is implicated as one of the initial triggers for pulmonary vasoconstriction and vascular remodeling in patients with PAH and animals with experimental PH. Loss of vasodilative effect of the mechanosensitive Ca²⁺ influx through TRPC6 and Piezo1 in PAECs, when endothelial injury and inflammation occur, would enhance vasoconstrictive effect of the mechanosensitive Ca²⁺ influx through TRPC6 and Piezo1 in PASMCs, thereby causing PH. TRPC6 and Piezo1 channels are nonselective cation channels, although the Ca²⁺ permeability (P_Ca) is greater than the Na⁺ permeability (P_Na) for both channels. The Na⁺ influx through TRPC6 and Piezo can also increase local Na⁺ concentration ([Na⁺]), which is a major driving force to induce inward transportation of Ca²⁺ through Na⁺/Ca²⁺ exchangers (64–68); the late phase of Ca²⁺ inward transportation via Na⁺/Ca²⁺ may contribute to the maintaining of high levels of [Ca²⁺]_cyt in mechanically stimulated PAECs and PASMCs to induce vasodilation and vasoconstriction, respectively.

Functional TRPC6 channels are homo- or heterotetramers composed of either TRPC6 subunits (homotetrameric channel) or TRPC6 with other TRPC subunits (e.g., TRPC1, TRPC3, and TRPC7) (heterotetrameric channel) (69). Furthermore, TRPC6/TRPC1 dimers can also form functional tetramers with TRPC4 and TRPC5 (70). Topology of TRPC6 shows that there are six transmembrane domains (TM) with the predicted pore region between TM5 and TM6. The cytoplasmic N- and C-termini contain multiple phosphorylation sites (71), two glycosylation sites, Ca²⁺/CaM and inositol (1,4,5)-trisphosphate (IP₃) receptor phosphoinositide-binding site (CIRPIB), and a diacylglycerol-sensitive lipid trafficking domain (26, 30, 37, 69). As a nonselective cation channel, TRPC6 is more permeable to Ca²⁺ than to other cations (e.g., P_Ca/P_Na = 5; P_Na/P_Cs = 0.7) (26). As an important subunit for forming the receptor-operated Ca²⁺ channel, TRPC6 can be activated by diacylglycerol (26, 31, 72–74), as shown in this study in TRPC6-transfected HEK cells and human PAECs, and by IP₃ receptor (37, 75) and tyrosine phosphorylation (30, 33, 36). In addition to its sensitivity to receptor-operated mechanisms, TRPC6 can be activated by mechanical stimulation (53, 73), as shown in this study in TRPC6-transfected HEK cells and human PAECs.

The pulmonary vascular endothelium is always exposed to high level of ligands (e.g., vasoactive substances, mitogenic and angiogenic factors, and inflammatory cytokines) in the blood (or locally released from ECs and infiltrated inflammatory cells). Synergistic activation of TRPC6 in lung ECs by receptor-mediated second messengers (e.g., diacylglycerol) and mechanical stimulation enables the channel to simultaneously sense chemical and mechanical stimuli and regulate pulmonary vascular function. TRPC6 channels can be blocked by 2-APB (2-aminoethoxydiphenyl borate), a nonselective inhibitor of many subtypes of TRP channels (76), and by BI-749327, a selective small-molecule inhibitor of TRPC6 (48). The IC₅₀ of BI-749327 for human and animal TRPC6 is 13–19 nM, which is ∼85-fold less than the IC₅₀ for TRPC3 and 42-fold less than the IC₅₀ for TRPC7 (48). BI-749327 is thus much more selective for TRPC6 than other TRPC channels that can form heterotetrameric channels with TRPC6.

As shown in this study, OAG, a membrane-permeable analog of diacylglycerol, significantly enhanced the cation currents in human PASMCs (I_Cation) (Fig. 1C) and TRPC6-transfected HEK cells (I_TRPC6) (Fig. 3D); BI-749327 (50 nM) reversibly reduced I_Cation in human PASMCs (Fig. 1D) and I_TRPC6 in TRPC6-transfected HEK cells (Fig. 4). BI-749327 at 50 nM was also sufficient to inhibit OAG-mediated increases in [Ca²⁺]_cyt in human PAECs (Fig. 5). These data indicate that, in human PAECs, TRPC6 is an important channel contributing to the whole cell cation currents responsible for agonist- and stretch-induced increase in [Ca²⁺]_cyt in human PAECs. It is, however, unknown whether TRPC6 channels are mainly present in human lung ECs in homotetrameric channels or heterotetrameric channels (with other TRPC channel subunits).

Piezo1 is a newly discovered mechanosensitive channel (38, 43, 77). Functional Piezo1 channels exhibit a unique propeller-shaped, three-bladed homotrimer structure; each subunit contains 38 transmembrane (TM) helices, a C-terminal extracellular domain (CED), an anchor between TM36 and TM37, and a long (9 nm) intracellular beam between TM28 and TM29 (78–80). Single-channel conductance of Piezo1 is reported to be 41.3 ± 0.9. pS at negative potential (−120 to −40 mV) and 27.1 ± 1.2 pS at positive potential (+20 to +110 mV), which indicates a significant outward rectification at positive potential (81). Whole cell outward currents at positive potential are also much greater than the inward currents at negative potential. The outward rectification of Piezo1 channels is also shown in this study (Fig. 8). In addition to mechanical forces, Yoda1 is selectively activated by the small-molecule allosteric activator, Yoda1 (52, 82, 83). The allosteric binding pocket of Yoda1 is located in the mechanosensory domain of Piezo1 nearby (4-nm away) the central pore. Yoda1 functions as a molecular wedge facilitating force-induced conformational changes to lower Piezo1’s mechanical threshold for activation (52). In human PAECs, 5 µM Yoda1 induced an increase in [Ca²⁺]_cyt that was similar to membrane stretch-mediated increase in [Ca²⁺]_cyt (Fig. 9) indicating that Piezo1 is an important channel in sensing mechanical force during vascular extension and recruitment.

There is no selective blocker for Piezo1 channels although many cation channel blockers can inhibit cation currents through mechanosensitive channels. GxMTx4 is a spider venom peptide that has been shown to block Piezo1 channels (49) and inhibit stretch-activated cation channels in astrocytes, cardiomyocytes, smooth muscle cells, and skeletal muscle cells. In this study, we found that 5 µM GsMTx4 resulted in an 50%–55% inhibition of membrane stretch-mediated increases in [Ca²⁺]_cyt (54.9 ± 9.5%) and Yoda-1-mediated increases in [Ca²⁺]_cyt (55.3 ± 11.0%) in human PAECs (Fig. 7). GsMTx4 (5 µM) also significantly inhibited whole cell cation currents in human PAECs bathed in Hypo-Osm solution; extracellular application of GsMTx4 resulted in 35.8 ± 10.7% inhibition of the inward currents at −100 mV and 62/2 ± 7.0% inhibition of the outward currents at +100 mV (Fig. 6). In Piezo1-transfected HEK cells, we found that OAG, which significantly enhanced cation currents through TRPC6 channels (I_TRPC6) in TRPC6-transfected HEK cells, had no effect on cation currents through Piezo1 channels (I_Piezo1) (Fig. 8C). These results indicate that GsMTx4-sensitive cation currents through Piezo1 channels are important mechanosensitive currents that are insensitive to OAG in human PAECs.

In summary, mechanosensitive cation channels in ECs play a significant role in sensing mechanical forces including flow-mediated shear stretch and pressure-induced stretch in the lung vasculature. This study indicates that TRPC6 and Piezo1 are two important mechanosensitive cation channels present in human PAECs (Fig. 10). TRPC6, but not Piezo1, is also sensitive to diacylglycerol, an important second messenger that mediates receptor-operated signal transduction in many types of cells. It is thus critical to conduct further studies to define potential pathogenic roles of these two channels in human PAECs and whether and how to target these channels for developing future therapeutic approaches for pulmonary vascular disease.

GRANTS

This work is supported in part by the National Heart, Lung, and Blood Institute of the National Institutes of Health Grants HL135807 and HL146764.

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

A.M. and J.X.-J.Y. conceived and designed research; T.Z. performed experiments; T.Z., S.P., Z.S., M.W., L.Y., P.A.T., A.M., and J.X.-J.Y. analyzed data; T.Z., S.P., Z.S., M.W., L.Y., P.A.T., A.M., and J.X.-J.Y. interpreted results of experiments; T.Z., S.P., Z.S., M.W., L.Y., P.A.T., A.M., and J.X.-J.Y. prepared figures; A.M. and J.X.-J.Y. drafted manuscript; T.Z., S.P., Z.S., M.W., L.Y., P.A.T., A.M., and J.X.J.Y. edited and revised manuscript; T.Z., S.P., Z.S., M.W., L.Y., P.A.T., A.M., and J.X.-J.Y. approved final version of manuscript.