PASMCs (passages 5–10) from normal subjects (Lonza, Basel, Switzerland) and IPAH patients (kindly offered by Prof. Jason X.-J. Yuan) (18, 19) were cultivated in Medium 199 supplemented with fetal bovine serum (FBS, 10%; Thermo Fisher Scientific, Waltham, MA, USA), D-valine (50 μg/ml; MilliporeSigma, Burlington, MA, USA), endothelial cell growth supplement (20 μg/ml; BD Biosciences, Franklin Lakes, NJ, USA), penicillin G (100 U/ml), and streptomycin (100 μg/ml; Fujifilm Wako Pure Chemical, Osaka, Japan) at 37°C.

Animal experiments were approved by the Ethics Committees of Nagoya City University (H30-P-1) and Aichi Medical University (2019–15). To generate MCT-PH rats (20–22), rats (Sprague-Dawley, male, 4 weeks old, 100–110 g; Japan SLC, Hamamatsu, Japan) were subcutaneously injected with vehicle (saline) or MCT (60 mg/kg; MilliporeSigma) and bred for 3 weeks (23). To obtain SuHx-PH rats (20–22), rats (Sprague-Dawley, male, 6 weeks old, 220–250 g; Japan SLC) were subcutaneously injected with vehicle (saline) or Sugen 5416 (20 mg/kg; MedChemExpress, Monmouth Junction, NJ, USA) and bred in a hypoxic chamber (10% O₂) with a ProOx110 O₂ controller (Biospherix, Parish, NY, USA) for 3 weeks and thereafter in a normoxic chamber for 2 weeks. Regarding hypoxia-PH mice (20–22), mice (C57BL/6, male, 8 weeks old, 18–23 g; Japan SLC) were bred in a normoxic or hypoxic (10% O₂) chamber for 4 weeks. Rats/mice were anesthetized by an intraperitoneal injection of ketamine (100 mg/kg) and xylazine (26 mg/kg). The pulmonary artery (the first to third branches) and right ventricle were dissected in Ca²⁺/Mg²⁺-free Krebs solution (in mM; 112 NaCl, 4.7 KCl, 25 NaHCO₃, 1.2 KH₂PO₄, 14 glucose, and pH 7.4 by gassing with 95% O₂/5% CO₂). The endothelium layer of the pulmonary artery was stripped out by water flow.

Total RNA was extracted using RNAiso Plus (Takara Bio, Kusatsu, Japan) and reverse transcribed to cDNA using the ReverTra Ace qPCR RT Master Mix (Toyobo, Osaka, Japan) (24). A qPCR analysis was carried out using SYBR Premix Ex Taq (Takara Bio) by the LightCycler 96 qPCR system (Roche Diagnostics, Basel, Switzerland). Specific primers were designed as shown in Table 1.

Protein fraction was extracted using RIPA buffer (for human PASMCs) and T-PER Tissue Protein Extraction Reagent (for rat/mouse pulmonary arterial smooth muscles (PASMs); Thermo Fisher Scientific) (23). Extracted protein (20 μg/lane) was applied to an acrylamide gel (8%) and transferred to an Immobilon-P PVDF membrane (MilliporeSigma). The membrane was blocked with Tris-buffered saline containing bovine serum albumin (5%) and Tween 20 (0.1%; MilliporeSigma) at room temperature (25°C) for 3 h and then treated with a primary antibody for KCNK1 (1:800; APC-110), KCNK2 (1:800; APC-047), KCNK3 (1:800; APC-024), KCNK6 (1:800; APC-040, Alomone Labs, Jerusalem, Israel), JNK (1:1000; #9252), or phospho (p)-JNK (1:1000; #4668, Cell Signaling Technology, Danvers, MA, USA) at 4°C for 18 h. Immunoblotted membranes were then exposed to an anti-rabbit HRP-conjugated IgG secondary antibody (1:5000; #170-6515, Bio-Rad Laboratories, Hercules, CA, USA) at room temperature for 1 h. Blotting signals were detected using an ImmunoStar LD reagent (Fujifilm Wako Pure Chemical) and observed with the Imager 600 system (GE HealthCare Technologies, Chicago, IL, USA). Protein expression levels were normalized using anti-β-actin (1:5000; A5316, MilliporeSigma) and anti-mouse HRP-conjugated IgG (1:10000; #170-6516, Bio-Rad Laboratories) antibodies.

The lungs of MCT-PH rats were fixed with paraformaldehyde (4%; MilliporeSigma) in PBS. Paraffin-embedded sections from lung lobes were prepared by the Biopathology Institute (Oita, Japan). They were deparaffinized and heat-induced epitope retrieval was carried out with Tris-HCl (10 mM, pH 9.0) containing EDTA (1 mM) at 115°C for 10 min. As the first step, sections were treated with a KCNK1 or KCNK2 antibody (1:100) using an ImmPRESS HRP reagent kit (Vector Laboratories, Burlingame, CA, USA) at room temperature for 1 h. After washing twice in PBS, they were covered with a secondary antibody contained in the ImmPRESS HRP reagent kit at room temperature for 30 min and rinsed twice with PBS. They were then treated with Fluorescein (1:200) in 1×plus amplification diluent (Akoya Biosciences, Marlborough, MA, USA) at room temperature for 10 min. After heating in a microwave for 1 min and washing with PBS, the same sections were treated with an α-smooth muscle actin (α-SMA) antibody (1:1000; #19245, Cell Signaling Technology), the ImmPRESS HRP reagent kit, and Cyanine 3 (1:400) using the same protocol as the first step. They were also stained with 4’,6-diamidino-2-phenylindole (DAPI; Dojindo Laboratories, Kumamoto, Japan). Immunohistochemical images were obtained using the Aperio CS2 image capture device (Leica Biosystems, Wetzlar, Germany).

Human PASMCs (3 × 10³ cells/well) were seeded on a 96-well plate (Falcon #353075, Corning, Corning, NY, USA) and cultured at 37°C for 6 h (23). Thereafter, they were treated with culture medium containing FBS (10%) and the vehicle [dimethyl sulfoxide (DMSO)] or drug for 48 h. Cell viability was assessed using Cell Counting Kit-8 (Dojindo Laboratories) based on the WST-8 assay. Cell proliferation was assessed using the Cell Proliferation ELISA, BrdU kit (Roche Diagnostics).

Human PASMCs (5 × 10⁴ cells/well) were seeded on a 24-well Transwell insert with a membrane pore size of 8 μm (#3422, Corning) (23). Thereafter, they were treated with culture medium containing FBS (1%) in the upper chamber and that containing FBS (10%) and the vehicle (DMSO) or drug in the lower chamber for 24 h. Transwell inserts were fixed in paraformaldehyde (4%) and stained with crystal violet (1%; Fujifilm Wako Pure Chemical). The number of migratory cells was counted from digital images of Transwell inserts using the SMZ1270 stereomicroscope system equipped with a DS-Vi1 color microscope camera and NIS-Elements imaging software (Nikon, Tokyo, Japan).

Human PASMCs (3 × 10³ cells/well) were seeded on a 96-well plate (Falcon #353075, Corning) and cultured at 37°C for 6 h. Thereafter, they were transiently transfected with universal negative control, KCNK1 ((+) UUGCCAUGUUGGUAGUUCUdTdT and (-) AGAACUACCAACAUGGCAAdTdT), or KCNK2 ((+) GGAAACACCUCCAAUCAAAdTdT and (-) UUUGAUUGGAGGUGUUUCCdTdT) siRNA construct (20 nM; Nippon Gene, Tokyo, Japan) using Lipofectamine RNAiMax transfection reagent (Thermo Fisher Scientific). The culture medium was replaced with siRNA-free medium 12–24 h after transfection. Experiments using siRNA were performed 48 h after transfection.

Human PASMCs were incubated with the voltage-sensitive fluorescent dye, DiBAC₄(3) (100 nM; Dojindo Laboratories), at room temperature for 30 min. DiBAC₄(3) at the same concentration was added to the extracellular solution during measurements. Fluorescent signals were measured using the A1R confocal fluorescence imaging system equipped with an ECLIPSE Ti inverted microscope, a Plan Apo VC objective lens (20×/0.75), NIS-Elements imaging software (Nikon), and a solid-state 488-nm laser (Coherent, Santa Clara, CA, USA). PASMCs were illuminated at a 488-nm wavelength, and the fluorescent emissions (>520 nm) were obtained every 5 s. Membrane potential is presented as F/F_140K, where F is the fluorescence intensity and F_140K is the maximum fluorescence intensity in the 140-mM K⁺ HEPES-buffered solution (theoretically 0 mV). Standard HEPES-buffered solution was used as an extracellular solution (in mM): 137 NaCl, 5.9 KCl, 2.2 CaCl₂, 1.2 MgCl₂, 14 glucose, 10 HEPES, and pH 7.4 with NaOH. In the 140-mM K⁺ HEPES-buffered solution, the concentrations of NaCl and KCl in standard HEPES-buffered solution were changed to 2.9 and 140 mM, respectively.

Human PASMCs were loaded with fura-2 acetoxymethyl ester (fura-2/AM, 10 μM; Thermo Fisher Scientific) at room temperature for 30 min. [Ca²⁺]_cyt measurements were performed using the fluorescence imaging system equipped with an ECLIPSE Ti2 inverted microscope, a S FL objective lens (20×/0.75), NIS-Elements imaging software (Nikon), a pE-340^fura LED illuminator (CoolLED, Hampshire, UK), and a C9100-12 EM-CCD digital camera (Hamamatsu Photonics, Hamamatsu, Japan). PASMCs were illuminated at 340-/380-nm wavelengths, and the fluorescent emissions (510/80 nm) were obtained every 5 s. The fura-2 signal is presented as the fluorescence ratio (F₃₄₀/F₃₈₀). Standard HEPES-buffered solution was used as an extracellular solution.

Pharmacological reagents were obtained from Fujifilm Wako Pure Chemical, except for EDTA, HEPES (Dojindo Laboratories), and quinine (MilliporeSigma). Quinine and tetrapentylammonium (TPA) were dissolved in DMSO at concentrations of 150 and 100 mM, respectively, as a stock solution.

Pooled data are shown as the means ± S.E. The significance of differences between two groups was examined using the non-parametric Mann–Whitney U test (n < 10) or Student's t-test (n ≥ 10) using BellCurve software (Social Survey Research Information, Tokyo, Japan). The significance of differences among groups was assessed by Scheffé's or Steel's test after non-parametric Kruskal-Wallis test (n < 10) or Scheffé's test after an analysis of variance (ANOVA) (n ≥ 10) using the same software.

The expression of KCNK channel genes (KCNK1 to 18, except for 8, 11, and 14) was analyzed in PASMCs from normal subjects and IPAH patients by qPCR. KCNK2/TREK1, KCNK3/TASK1, and KCNK6/TWIK2 genes were detected in normal-PASMCs (Figure 1A and Supplementary Table S1). The mRNA expression of KCNK1/TWIK1 (0.00061 ± 0.00012 of β-actin, n = 4, p = 0.029 vs. normal, 0.00002 ± 0.00001, n = 4; IPAH/normal ratio = 34.70 ± 7.05-fold) and KCNK2 (0.00319 ± 0.00054, n = 4, p = 0.029 vs. normal, 0.00154 ± 0.00014, n = 4; 2.07 ± 0.35-fold) channels was up-regulated in IPAH-PASMCs. In contrast, the mRNA expression of KCNK3 (0.00006 ± 0.00004, n = 4, p = 0.029 vs. normal, 0.00025 ± 0.00003, n = 4; 0.26 ± 0.14-fold) and KCNK6 (0.00366 ± 0.00034, n = 4, p = 0.029 vs. normal, 0.01185 ± 0.00083, n = 4; 0.31 ± 0.03-fold) channels was down-regulated in IPAH-PASMCs.

The expression levels of the KCNK1, 2, 3, and 6 proteins in normal- and IPAH-PASMCs were compared by Western blotting. The protein expression of KCNK1 channels was higher in IPAH-PASMCs than in normal-PASMCs (1.51 ± 0.17-fold, n = 6, p = 0.041 vs. normal, 1.00 ± 0.14, n = 6) (Figure 1B). The protein expression of KCNK2 was also higher in IPAH-PASMCs (1.29 ± 0.05-fold, n = 7, p = 0.002 vs. normal, 1.00 ± 0.04, n = 7). In contrast, the protein expression levels of KCNK3 and KCNK6 were lower in IPAH-PASMCs than in normal-PASMCs (0.82 ± 0.05-fold, n = 8, p = 0.015 and 0.79 ± 0.04-fold, n = 8, p = 0.028, respectively). These results indicated that the expression of KCNK1 and KCNK2 channels was up-regulated, whereas that of KCNK3 and KCNK6 channels was down-regulated in PASMCs from IPAH patients.

Since the expression of KCNK1 and KCNK2 channels was up-regulated in IPAH-PASMCs, their expression changes in PASMs from three types of experimental PH animals (MCT-PH rats, SuHx-PH rats, and hypoxia-PH mice) were also examined. It was confirmed that endothelium was not attached to the dissected PASMs because endothelium-dependent relaxation was not induced by acetylcholine (25). The expression of KCNK1 and KCNK2 channel proteins was examined in PASMs from control and MCT-PH rats by Western blotting. The protein expression of KCNK1 (1.72 ± 0.10-fold, n = 8, p < 0.001 vs. control, 1.00 ± 0.09, n = 8) and KCNK2 (1.35 ± 0.06-fold, n = 8, p = 0.002 vs. control, 1.00 ± 0.06, n = 8) channels was increased in MCT-PASMs (Figures 2A,B). In addition, the expression of KCNK1 and KCNK2 channel proteins was analyzed using the lung sections of control and MCT-PH rats by immunohistochemical staining. Immunohistochemical images revealed that KCNK1 and KCNK2 channels were localized in the medial (smooth muscle) layer of the pulmonary artery and their expression was higher in MCT-PH rats than in the control rats (Figures 2C,D).

Changes in KCNK1 and KCNK2 protein expression were then assessed in PASMs from control and SuHx-PH rats by Western blotting. The expression of KCNK1 (1.40 ± 0.13-fold, n = 6, p = 0.002 vs. control, 1.00 ± 0.04, n = 6) and KCNK2 (1.20 ± 0.06, n = 6, p = 0.041 vs. control, 1.00 ± 0.06, n = 6) channels was up-regulated in PASMs from SuHx-PH rats (Figures 3A,B).

KCNK1 and KCNK2 protein expression levels were also evaluated in PASMs from normoxia and hypoxia-PH mice. The expression of KCNK1 (2.14 ± 0.12-fold, n = 6, p = 0.002 vs. normoxia, 1.00 ± 0.06, n = 6) and KCNK2 (1.41 ± 0.04-fold, n = 6, p = 0.002 vs. normoxia, 1.00 ± 0.06, n = 6) channels was increased in PASMs from hypoxia-PH mice (Figures 3C,D). These results strongly suggest the up-regulated expression of KCNK1 and KCNK2 channels in PASMs from three types of experimental PH animals, similar to PAH patients.

Since vascular remodeling in PAH is predominantly mediated by the excessive proliferation of PASMCs (4, 5), the involvement of KCNK channels was examined using Cell Counting Kit-8. Cell viability increases in a time-dependent manner in both normal-PASMCs (A₄₅₀ = 0.393 ± 0.028 at 0 h, 0.775 ± 0.020 at 24 h, 0.906 ± 0.052 at 48 h, and 1.004 ± 0.067 at 72 h, n = 4) and IPAH-PASMCs (0.441 ± 0.008 at 0 h, 0.909 ± 0.019 at 24 h, 1.184 ± 0.042 at 48 h, and 1.330 ± 0.067 at 72 h, n = 5) (Figure 4A). The growth of IPAH-PASMCs at 48 and 72 h was markedly greater than that of normal-PASMCs (p = 0.020 and p = 0.004, respectively), which is consistent with previous findings (23, 26, 27).

The effects of KCNK channel blockers on the facilitated proliferation of IPAH-PASMCs were examined. In IPAH-PASMCs, increases in proliferation were reduced by a treatment with 300 μM quinine, a KCNK channel blocker (10, 28, 29), for 48 h (A₄₅₀ = 0.552 ± 0.032, n = 5, p = 0.008 vs. control, 1.057 ± 0.032, n = 5) (Figure 4B). A similar reduction was observed with 100 μM TPA, another KCNK channel blocker that is structurally different from quinine (29–32), for 48 h (0.703 ± 0.039, n = 5, p = 0.008 vs. control, 1.030 ± 0.013, n = 5) (Figure 4C). This result suggests that up-regulated KCNK channel function involved in the enhanced proliferation of IPAH-PASMCs.

The effects of KCNK channel blockers on the migration of IPAH-PASMCs were investigated by Transwell assays. The migration was inhibited by 300 μM quinine for 24 h (2,065 ± 76 cells, n = 4, p = 0.039 vs. control, 2,720 ± 88 cells, n = 4) (Figure 5). It was also reduced by 100 μM TPA for 24 h (1,710 ± 72 cells, n = 4, p = 0.039). Collectively, these results suggest that up-regulated KCNK channels mediated the migration and excessive proliferation of IPAH-PASMCs.

To obtain direct evidence for the contribution of KCNK1 and KCNK2 channels in the proliferation of IPAH-PASMCs, cell viability and proliferation were examined by WST-8 and BrdU incorporation assays after KCNK1 or KCNK2 specific knockdown by siRNAs. The knockdown efficacy by KCNK1 and KCNK2 siRNA was confirmed by qPCR and Western blotting. KCNK1 siRNA knocked-down the mRNA expression of KCNK1 channels in IPAH-PASMCs (92.4 ± 5.0% decrease, n = 6, p = 0.002 vs. control siRNA, n = 6), whereas it did not affect the expression levels of KCNK2, KCNK3, or KCNK6 channels (n = 6, p > 0.05) (Figure 6A). KCNK2 siRNA knocked-down KCNK2 expression (78.1 ± 2.2% decrease, n = 6, p = 0.002 vs. control siRNA, n = 6), but did not affect the expression of KCNK1, KCNK3, or KCNK6 channels (n = 6, p > 0.05) (Figure 6B). Similarly, KCNK1 siRNA knocked-down KCNK1 proteins (66.3 ± 4.4% decrease, n = 6, p = 0.002) (Figure 6C) and KCNK2 siRNA knocked-down KCNK2 proteins (43.0 ± 6.3% decrease, n = 6, p = 0.002) (Figure 6D) in IPAH-PASMCs.

When KCNK1 siRNA was transfected into IPAH-PASMCs for 48 h, their excessive proliferation was significantly attenuated (9.3 ± 2.3% decrease, n = 19, p = 0.025 vs. control siRNA, n = 19) (Figure 7A). KCNK2 siRNA for 48 h also suppressed the facilitated proliferation (11.6 ± 2.0% decrease, n = 19, p = 0.013 vs. control siRNA, n = 19) (Figure 7B). To confirm the results obtained from the WST-8 assay, the BrdU assay was performed on IPAH-PASMCs. The excessive proliferation was inhibited by the siRNA knockdown of KCNK1 (20.5 ± 1.8% decrease, n = 28, p < 0.001 vs. control siRNA, n = 28) or KCNK2 (33.5 ± 2.0% decrease, n = 28, p < 0.001 vs. control siRNA, n = 28) for 48 h (Figures 7C,D).

The effects of KCNK1 and KCNK2 siRNA on the migration of IPAH-PASMCs were assayed using the Transwell plates. KCNK1 siRNA inhibited the migration of IPAH-PASMCs for 24 h (1,394 ± 47 cells, n = 4, p = 0.039 vs. control siRNA, 2,189 ± 139 cells, n = 4) (Figure 7E). KCNK2 siRNA also suppressed the migration of IPAH-PASMCs for 24 h (1,130 ± 20 cells, n = 4, p = 0.039). Collectively, these findings strongly indicate that the activities of KCNK1 and KCNK2 channels were responsible for the proliferation and migration of IPAH-PASMCs, leading to vascular remodeling in PAH.

We examined whether the activities of KCNK1 and KCNK2 channels contributed to the resting membrane potential of IPAH-PASMCs by fluorescence imaging using the voltage-sensitive dye, DiBAC₄(3). In IPAH-PASMCs, the resting membrane potential was shifted in the depolarizing direction by the siRNA knockdown of KCNK1 channels (F/F_140K = 0.617 ± 0.007, n = 131, p < 0.001 vs. control siRNA, 0.253 ± 0.007, n = 115) (Figures 8A,B). Similar depolarizing changes in the resting membrane potential were induced by the siRNA knockdown of KCNK2 channels (0.582 ± 0.008, n = 98, p < 0.001).

Since membrane depolarization decreases [Ca²⁺]_cyt in a proliferative phenotype of vascular myocytes (33, 34), the effects of the siRNA knockdown of KCNK1 and KCNK2 channels on resting [Ca²⁺]_cyt in IPAH-PASMCs were examined by fluorescence imaging using the Ca²⁺ indicator, fura-2/AM. KCNK1 siRNA-treated IPAH-PASMCs showed lower resting [Ca²⁺]_cyt than control siRNA-treated cells (F₃₄₀/F₃₈₀ = 0.255 ± 0.017, n = 63, p < 0.001 vs. control siRNA, 0.367 ± 0.018, n = 73) (Figures 8C,D). KCNK2 siRNA-treated cells also had a lower resting [Ca²⁺]_cyt than control cells (0.216 ± 0.005, n = 48, p < 0.001). Taken together, these findings suggest that up-regulated KCNK1 and KCNK2 channel expression caused membrane hyperpolarization and subsequent [Ca²⁺]_cyt increases in IPAH-PASMCs.

To elucidate the involvement of KCNK1 and KCNK2 channels in the proliferation and migration of IPAH-PASMCs, we focused on the JNK signaling pathway, which is involved in vascular remodeling in PAH (12–17). Western blot analyses revealed that the phosphorylation level of JNK was facilitated in IPAH-PASMCs compared to in normal-PASMCs (1.65 ± 0.04-fold, n = 6, p = 0.002 vs. normal, 1.00 ± 0.06, n = 6) (Figure 9A). Up-regulated JNK phosphorylation in IPAH-PASMCs was decreased by the siRNA knockdown of KCNK1 channels (43.7 ± 3.3% decrease, n = 6, p = 0.002) (Figure 9B). Similarly, it was down-regulated by the siRNA knockdown of KCNK2 channels (48.9 ± 2.7% decrease, n = 6, p = 0.002) (Figure 9C). These findings suggest that the expression and activity of KCNK1 and KCNK2 channels affected the phosphorylation levels of JNK in IPAH-PASMCs, thereby facilitating the proliferation and migration of PASMCs in PAH.