Omega-3 fatty acid epoxides produced by PAF-AH2 in mast cells regulate pulmonary vascular remodeling

Pulmonary arterial hypertension (PAH) is a rare, fatal disease characterized by elevated pulmonary artery pressure and progressive right heart failure due to pulmonary vascular remodeling involving endothelial cells, smooth muscle cells, and fibroblasts. Although current vasodilator therapies improve outcomes, many patients have inadequate responses, underscoring the need for disease-modifying treatments that suppress vascular remodeling. Inflammatory cells in lung tissue release cytokines, chemokines, and lipid mediators that influence inflammation, thrombosis, angiogenesis, and fibrosis, thereby modulating remodeling. Proinflammatory lipids (e.g., prostanoids, leukotrienes) contribute to PH pathogenesis, whereas ω-3 fatty acids (EPA, DHA) and some derivatives have bioprotective, anti-remodeling effects. Prior work showed an ω-3 metabolite (18-HEPE) from macrophages restrains cardiac fibroblast activation in pressure overload. The authors hypothesized that ω-3 fatty acid-derived epoxides produced in the lung could suppress pulmonary vascular remodeling in PH. They conducted lipidomics and genetic/pharmacologic studies to test whether ω-3 epoxides generated by PAF-AH2 in mast cells regulate fibroblast activation and PH progression, and whether human PAF-AH2 variants are implicated in PAH.

Background literature indicates: (1) Pulmonary vascular remodeling drives PH and involves multiple vascular and inflammatory cell types; (2) lipid mediators regulate inflammation, thrombosis, angiogenesis, and fibrosis in vascular disease; (3) proinflammatory lipids like leukotrienes contribute to PH; (4) ω-3 fatty acids and specialized derivatives have cardioprotective and anti-remodeling actions; (5) ω-3 epoxides can have potent anti-inflammatory and vasodilatory effects; and (6) PAF-AH2 in mast cells can generate ω-3 epoxides and modulate mast cell activation. These data motivated investigation of the PAF-AH2–ω-3 epoxide axis in PH.

Study approval: Animal procedures conformed to NIH guidelines and were approved by Keio University. Human genetics was IRB-approved with consent.

Animal models and genotypes: Pafah2 knockout (KO) mice (backcrossed >10 generations to C57BL/6J), Pla2g7 KO mice (plasma-type PAF-AH KO), and mast cell-deficient KitW-sh/W-sh mice (C57BL/6J background). Male mice 7–10 weeks were used (sex effects assessed in Supplementary Fig. 3).

PH induction: (1) Chronic hypoxia: mice housed at 10% O2 for 4 or 8 weeks. (2) Sugen/hypoxia (SuHx): SU5416 (VEGF inhibitor) 20 mg/kg subcutaneously on days 0, 7, 14, with 10% O2 housing for 7 weeks.

Lipidomics: LC/ESI-MS/MS quantified EPA/DHA/AA metabolites from right lung tissue. Lipids were extracted (Bligh and Dyer), purified by solid-phase extraction (InertSep NH2), and analyzed on Shimadzu UPLC with AB SCIEX QTRAP 4500. Targeted MRM transitions were used. Z-scores computed across groups (n=3/timepoint in Fig. 1 heatmaps). Western blots assessed PAF-AH2 protein in lung homogenates.

Hemodynamics and histology: Right ventricular systolic pressure (RVSP) by Millar catheter via right jugular vein under isoflurane. RV hypertrophy quantified as RV/(LV+septum). Lungs fixed and paraffin-embedded for H&E, Elastica-van Gieson (EVG), toluidine blue, and immunostaining (CD31, Vimentin; PAF-AH2; Tryptase). Pulmonary arteriole wall thickness was measured on vessels <100 μm: [(outer diameter − lumen diameter)/outer diameter] × 100.

Cellular sources and adoptive transfer: Immunostaining localized PAF-AH2 to tryptase+ mast cells in mouse PH models (hypoxia, SuHx) and human idiopathic PAH lungs. For functional testing, KitW-sh/W-sh mice were reconstituted intravenously with 5×10^6 WT or Pafah2 KO bone marrow–derived mast cells (BMMCs). Four weeks later, mice were exposed to hypoxia for 4 weeks and evaluated (RVSP, RV/LV+S, wall thickness, lung Col1a1 mRNA, RV Nppa).

Mast cell assays: BMMCs generated in IL-3–containing medium (4–6 weeks; >95% Kit+ FcεRI+). Degranulation assessed by β-hexosaminidase release. Hypoxia-induced degranulation compared between WT and Pafah2 KO BMMCs; in vivo degranulation modulated by Ketotifen (mast cell stabilizer) administered in drinking water for 4 weeks during hypoxia.

Fibroblast assays: Primary murine lung fibroblasts isolated via collagenase digestion and culture. Lipid extracts (free fatty acid fraction) from BMMC-conditioned media (Sep-Pak C18 fractionation) were applied to fibroblasts. Proliferation (RealTime-Glo MT), migration (Boyden chamber), myofibroblast markers (SM22α immunostaining), and mRNA (Col1a1, Acta2, Il6 by qPCR) were measured. Direct treatments included 17,18-EpETE, 19,20-EpDPE, their dihydrodiols, EPA, DHA, and ω-6 epoxide 14,15-EET (typically 1 μM). TGF-β stimulation (1.0–2.5 ng/ml) tested pathway effects; p-Smad2 and total Smad2 analyzed by Western blot.

Endothelial and smooth muscle cell assays: Human pulmonary artery endothelial cells (HPAECs) and smooth muscle cells (PASMCs) tested for gene expression changes, oxidative stress protection (ROS assay with H2O2), viability, and proliferation in response to ω-3 epoxides and BMMC lipid extracts.

Regulation of Pafah2: BMMCs cultured under hypoxia in vitro; HIF activation by DMOG or CoCl2 to assess Pafah2 expression; Cyp4a12 (ω-3 epoxidation enzyme) expression measured.

In vivo ω-3 epoxide therapy: 19,20-EpDPE administered intraperitoneally at 0.05 mg/kg/day starting 2 weeks after hypoxic exposure in hypoxia model and 3 weeks after hypoxic exposure in SuHx model. Comparators included PBS/vehicle and DHA (0.05 mg/kg/day). Outcomes: arteriole wall thickness, RVSP, RV/(LV+S), and perivascular fibrosis (Vimentin immunostaining).

Human genetics: Whole-exome sequencing in 262 PAH patients (90 idiopathic, 61 heritable, 54 CTD-associated; mean age 45.5±16.6 years; 78% female; mean PAP 42.9±15.5 mmHg). Variants in PAFAH2 prioritized with CADD≥30 and predicted deleteriousness (SIFT, PolyPhen-2). Known PAH genes also annotated.

Variant functional studies and modeling: Human PAFAH2 cDNA cloned into pcDNA3.1(+); site-directed mutagenesis created variants R85C (c.253C>T), Q184R (c.551A>G), and catalytic-site S236C (c.707C>G). HEK293 cells transfected; protein levels assessed by Western blot with/without proteasome inhibitor MG132 (10 μM, 6 h). Homology model based on plasma PAF-AH (PDB 5i9i; 42.3% identity) and molecular dynamics simulations (NAMD, CHARMM36, 72 ns) evaluated conformational impacts of variants.

Statistics: Data mean±SEM; two-tailed Student’s t test for two-group comparisons; one-way/two-way ANOVA with appropriate post hoc tests for multiple groups; survival by log-rank test; significance at P<0.05.

• Lipidomics in hypoxic PH mouse lungs showed significant reductions of ω-3 epoxides and their dihydrodiols over time under 10% O2: 17,18-EpETE (P=0.0033), 19,20-EpDPE (P=0.0455), 17,18-diHETE (P=0.026), 19,20-diHDOPE (P=0.0252). PAF-AH2 protein expression was also reduced in hypoxic lungs.
• Pafah2 KO mice under chronic hypoxia (8 weeks) had lower pulmonary ω-3 epoxide levels (17,18-EpETE P=0.0102; 19,20-EpDPE P=0.0070) than WT and exhibited exacerbated PH: increased pulmonary arteriole wall thickness (P<0.0001), higher RVSP (P=0.0098), greater RV hypertrophy (RV/(LV+S) P<0.0001), elevated RV Nppa mRNA (P=0.0126), and markedly worse survival (nearly 80% mortality by day 100; log-rank WT vs KO P=0.002). Pla2g7 KO mice did not show significant worsening vs WT in these PH measures.
• Hypoxic Pafah2 KO lungs showed pronounced perivascular fibrosis: increased lung Col1a1 mRNA and Vimentin+ area around vessels.
• PAF-AH2 expression localized predominantly to tryptase+ mast cells in mouse PH models (hypoxia and SuHx) and human idiopathic PAH lungs; mast cells accumulated perivascularly in PH.
• Mast cell-deficient KitW-sh/W-sh mice reconstituted with Pafah2 KO BMMCs developed more severe hypoxic PH than those reconstituted with WT BMMCs: increased wall thickness (P<0.0001), higher RVSP (P=0.0135), greater RV/(LV+S) (P=0.0044), higher RV Nppa (P=0.0431), and higher lung Col1a1 (P=0.0441), implicating mast cell PAF-AH2 in protection.
• Mast cell degranulation was modestly increased by hypoxia but did not differ between WT and Pafah2 KO in vivo or in vitro; Ketotifen suppressed degranulation but did not ameliorate severe PH in Pafah2 KO mice, indicating degranulation-independent mechanisms.
• Lipid extracts (free fatty acid fraction) from Pafah2 KO BMMC supernatants enhanced lung fibroblast proliferation and activation marker expression (Col1a1, Acta2), which were suppressed by adding ω-3 epoxides (17,18-EpETE or 19,20-EpDPE). Direct treatment with ω-3 epoxides (1 μM) reduced fibroblast proliferation and migration and inhibited TGF-β–induced upregulation of Col1a1, Acta2, and Il6 and SM22α expression; EPA/DHA or dihydrodiols were ineffective; ω-6 epoxide 14,15-EET did not reproduce the inhibitory effects. ω-3 epoxides attenuated TGF-β signaling as shown by reduced Smad2 phosphorylation (P=0.0284 at 1.0 and 2.5 ng/ml TGF-β).
• ω-3 epoxides did not substantially alter PA endothelial cell gene expression, antioxidant responses, or PASMC proliferation in vitro under tested conditions.
• Hypoxia/HIF activation downregulated Pafah2 expression in BMMCs in vitro; Cyp4a12 expression was unchanged.
• Therapeutic administration of 19,20-EpDPE (0.05 mg/kg/day i.p.) starting 2 weeks after hypoxia significantly improved PH in both WT and Pafah2 KO mice, reducing wall thickness, RVSP, RV/(LV+S), and perivascular fibrosis; DHA or 14,15-EET did not show benefit. In SuHx mice, 19,20-EpDPE (but not DHA) also reduced wall thickness (P=0.0002–0.0003), RVSP (P=0.0048–0.0341), and RV/(LV+S) (P=0.0035–0.0057).
• Whole-exome sequencing in 262 PAH patients identified two rare PAFAH2 missense variants with high predicted pathogenicity: p.R85C and p.Q184R (distinct from catalytic residues). Molecular dynamics suggested conformational differences vs native protein. In HEK293 cells, mutant proteins had significantly reduced expression, partially rescued by proteasome inhibition (MG132), indicating enhanced proteasomal degradation; a catalytic-site variant (S236C) did not show reduced expression. These variants may predispose to PH by reducing PAF-AH2 protein stability and ω-3 epoxide production.

This study demonstrates that ω-3 epoxides generated by PAF-AH2 in lung mast cells suppress pulmonary vascular remodeling and disease progression in PH. In hypoxic PH, endogenous ω-3 epoxides and PAF-AH2 expression decline, and genetic loss of Pafah2 exacerbates PH, particularly perivascular fibrosis and right heart strain. Cell localization and adoptive transfer pinpoint mast cells as the protective source of ω-3 epoxides. Mechanistically, ω-3 epoxides act on perivascular fibroblasts to reduce proliferation, migration, and myofibroblast activation by attenuating TGF-β/Smad2 signaling, without evident direct effects on endothelial or smooth muscle cells under the conditions tested. The exacerbated phenotype in Pafah2 KO mice was not explained by differences in mast cell degranulation, and stabilizing mast cells with Ketotifen did not improve PH, supporting a degranulation-independent, paracrine lipid-mediator mechanism. Therapeutically, exogenous supplementation with 19,20-EpDPE ameliorated vascular remodeling and hemodynamics in both chronic hypoxia and SuHx models, whereas DHA or ω-6 epoxide lacked efficacy at the tested dose, indicating specific bioactivity of ω-3 epoxides. The absence of a similar phenotype in Pla2g7 KO mice and the known unique substrate specificity of PAF-AH2 underscore the particular importance of the PAF-AH2–ω-3 epoxide pathway. Identification of likely pathogenic PAFAH2 variants in PAH patients, which destabilize PAF-AH2 protein via proteasomal degradation, links this pathway to human disease and suggests a subgroup who may benefit from ω-3 epoxide–based therapies. Collectively, these findings position mast cell–derived ω-3 epoxides as endogenous antifibrotic mediators that restrain pulmonary vascular remodeling and represent a promising disease-modifying therapeutic avenue in PH/PAH.

The work establishes a PAF-AH2–ω-3 epoxide axis in mast cells that limits pulmonary vascular remodeling by suppressing TGF-β–driven fibroblast activation. Loss of Pafah2 exacerbates PH, while exogenous 19,20-EpDPE mitigates disease in multiple models. Discovery of destabilizing PAFAH2 variants in PAH patients implicates impaired ω-3 epoxide production in human pathogenesis and highlights precision therapeutic potential. Future studies should (1) identify the precise cellular targets and receptors/signal transduction for ω-3 epoxides in fibroblasts; (2) evaluate pharmacologic strategies to augment ω-3 epoxide bioavailability (e.g., optimized delivery, soluble epoxide hydrolase inhibition) and dose-response; (3) develop and study knock-in mouse models of human PAFAH2 variants; and (4) translate findings into early-phase clinical trials, particularly for patients harboring PAFAH2 variants or with evidence of low ω-3 epoxides.

• Disease models: Chronic hypoxia primarily models group III PH; although SuHx provides a more PAH-like phenotype, findings may not fully generalize to all PAH etiologies.
• Mechanistic specificity: The exact receptors or binding partners through which ω-3 epoxides inhibit TGF-β/Smad2 signaling in fibroblasts were not identified.
• Cell-type scope: Limited direct effects were observed on endothelial cells and smooth muscle cells under tested conditions; broader conditions or chronic exposures were not explored.
• Genetic validation: Functional impact of human PAFAH2 variants was shown in heterologous HEK293 overexpression with proteasome inhibition; in vivo validation in knock-in models is pending.
• Mast cell biology: While degranulation was assessed, other mast cell functions (e.g., cytokine profiles) under hypoxia were not comprehensively characterized.
• Dosing/translation: Only one dose and route of 19,20-EpDPE was tested; long-term safety, pharmacokinetics, and optimal regimens require further study.
• Sample sizes: Some analyses had modest n, typical for preclinical studies; replication and larger cohorts would strengthen conclusions.