Endothelial cell Piezo1 mediates pressure-induced lung vascular hyperpermeability via disruption of adherens junctions

Significance Increased hydrostatic pressure in lung capillaries experienced during high altitude, head trauma, and left heart failure can lead to disruption of lung endothelial barrier and edema formation. We identified Piezo1 as a mechanical sensor responsible for endothelial barrier breakdown (barotrauma) secondary to reduced expression of the endothelial adherens junction proteins VE-cadherin, β-catenin, and p120-catenin. Endothelial-specific deletion or pharmacological inhibition of Piezo1 prevented lung capillary leakage, suggesting a therapeutic approach for preventing edema and associated lung failure.

pulmonary artery was cannulated rapidly; a second cannula was inserted into the left atrium for drainage of the pulmonary venous effluent. Lung and heart were removed en bloc from the thorax and suspended from a counter-weighted beam balance. Lung was perfused at constant flow (2 mL/min), temperature (37°C) and venous pressure (4 cmH2O). Microvascular filtration coefficient (Kf,c) was measured from the rate of lung wet weight gain after a step increase (10 or 16 cmH2O) in venous pressure for 20 minutes. The rate of weight gain was normalized by the lung dry weight and step size to calculate Kf,c in units of milliliters per minute per centimeters of water per gram. (4) Ex Vivo Albumin PS product measurement. Mouse lungs were perfused as described above without recirculation to measure albumin permeability × surface area (PS product), as previously described by us.(5) After a 20-min equilibration perfusion, measurements were made after a step increase (0, 10 or 16 cmH2O) in venous pressure for 5 minutes. Then, 125 I-labeled albumin (final activity of 60,000 cpm/mL, Anazao Health Corp, Tampa, FL) was infused into the perfusing liquid for 3 min via a side-arm in the pulmonary artery cannula. A sample of the venous effluent was collected. Vascular tracer was washed out for 6 min. The venous effluent sample and the entirety of lung tissue were counted for gamma-radioactivity. Albumin PS was computed in units of mL/min/g from A/ct, where A represents the lung counts (cpm/g dry lung), c, the venous effluent tracer concentration (cpm/mL), and t, the exposure time (min).
In Vivo Albumin PS product measurement Mice were anesthetized with an intraperitoneal injection of ketamine (100 mg/kg body weight) and xylazine (15 mg/kg body weight) after 24 hr sham or TAC surgery. 100 ul 125I-labeled albumin (10 uCi/mL) was injected into the left jugular vein. After 30 min post-injection, lungs were briefly flushed to eliminate vascular tracer and was counted for gamma-radioactivity. Blood sample was collected and radioactivity was counted for comparison. Albumin PS was computed in units of mL/min/g from A/ct, where A represents the lung counts (cpm/g dry lung), c, Blood concentration (cpm/mL), and t, the exposure time (min).

(5)
Transverse Aortic Constriction (TAC) Minimally invasive transverse aortic constriction was performed under 1.5% isoflurane mixed with 0.5 -1.0 L/min room air.(6) Under microscopic view, a 2 cm midline incision till the second rib was performed in the anterior neck. The sternum was subsequently cut down to the second rib level. The aortic arch was reached simply by separating the connective tissues and thymus. A 6-0 silk suture underneath the aortic arch between the innominate and left carotid arteries was tied against 25 G needle with a loose double knot. Then immediately the needle will be removed, the lungs re-inflated and the chest closed layer-by-layer with 6-0 silk suture. Implementation of outlined surgical procedure in its entirety was taken no more than 30 min. The success of TAC was confirmed by echocardiography, which determine the flow velocities ratio change from the left to right carotid arteries, the dimension of the blood vessels, and the left ventricular function. In the GsMTx4 group, 3 µM GsMTx4 (Tocris Bioscience) was IV administrated 30 minutes before the TAC surgery. Dosing of GsMTx4 based on previous studies with the peptide. Transmission Electron Microscopy Approximately 6-9 nm colloidal gold-albumin complexes were prepared as tracers to determine the route of protein leak across the blood gas barrier as previously described (8). Approximately 6-9 nm colloidal gold was prepared by heating separately 80 mL 0.35 mM HAuCl4 and 20 mL 0.2% trisodium citrate, 0.1% tannic acid and 2.5 mM K2CO3 to 60 C before mixing the two solutions together with stirring and then heating to a rapid boil for 20 min. The solution cooled for 1 h at 4 C before stabilizing with 125 µg/mL linoleic acid-oleic acid albumin (L9655, Sigma) and 200 µg/mL polyglutamic acid. Solutions were ultracentrifuged at 105,000 x g for 1.5 h at 4 C. Supernatant was discarded. Colloidal goldalbumin was then diluted to A520 = 1 in Hanks' Balanced Salt Solution (HBSS) with calcium and magnesium just before use.
Mice were anesthetized, and the pulmonary artery and left atrium were cannulated as described above, but without removed the heart and lungs. Flow rate was held between 2.0 to 2.2 mL/min for all animals. The vasculature was first precleared with HBSS with calcium and magnesium at. The pressure was elevated by 16 cmH2O for 15 min and the perfusate was simultaneously changed to the gold-albumin containing buffer. This was followed by a 3 min wash with HBSS. Mouse lungs were then fixed by ex vivo perfusion with 2.5% glutaraldehyde, 4% PFA, 0.1 M HEPES and 2 mM CaCl2. The pressure for washing and fixation steps remained elevated for experimental groups or kept at baseline for control groups.

Co-Immunoprecipitation
Monolayers of HLMVEC passage 4 to 5 were treated with 5 µM Yoda1 for 0, 15 or 30 min. Monolayers were washed in HBSS with Ca 2+ and Mg 2+ briefly, then monolayers were lysed in 500 µL/10 cm plate of a buffer consisting of 10 mM Tris/Cl pH 7.5,150 mM NaCl, 0.5 mM Ethylenediaminetetraacetic acid (EDTA) and 0.5% nonyl phenoxypolyethoxylethanol (NP-40) with protease inhibitors. Lysates were kept on ice for 30 min, pipetting up and down every 10 min before centrifugation at 16,000 x g for 15 min. For immunoprecipitation, 300 µg of protein of each sample were added to a tube with 1 µg VE-cadherin antibody (Abcam, ab33168) in a total volume of 500 µL and rotated end-over-end overnight at 4 C. The following day, 40 µL of protein G agarose bead slurry was added to each sample, and rotated end-over-end for 2 h before centrifugation at 2500 x g for 2 min. Supernatant was discarded and beads were washed 3 times in 500 µL wash buffer consisting of 10 mM Tris/Cl pH 7.5,150 mM NaCl and 0.5 mM EDTA with protease inhibitors. Protein was eluted from agarose beads by heating in 2x Laemmli buffer with 2-mercaptoenthanol for 10 min at 95 C. Blots were probed with anti-VE-cadherin (Santa Cruz, BV9) and anti-p120-catenin (Santa Cruz, sc1101).

Calpain activity
Calpain activity was analyzed by the calpain activity assay kit (Abcam, ab65308). The fluorometric assay detects cleavage of calpain substrate Ac-LLY-AFC. In brief, mice lungs were isolated following PBS perfusion, and homogenized in the supplied extraction buffer. Protein concentration was determined by BCA methods. Subsequently 100 µg of samples were loaded to a 96 well black coated polystyrene plate. After addition of reaction buffer and substrate, the plate was incubated for 60 min at 37 ºC. After incubation the fluorescence units of each well was measured with excitation at 400 nm and emission at 505 nm. Signals were then normalized to the sample protein concentration. Data were shown as the relative calpain activity.

Statistics
Statistical analysis was completed using GraphPad Prism7. One-way ANOVA corrected for multiple comparisons with Tukey's test was used for experiments with more than 2 groups. Student's t-test was used for experiments with two experimental groups.