Germline and Somatic Mutations in DNA Methyltransferase 3A (DNMT3A) Predispose to Pulmonary Arterial Hypertension (PAH) in Humans and Mice: Implications for Associated PAH

Background: Mutations are found in 10-20% of idiopathic PAH (IPAH) patients, but none are identified in connective tissue disease-associated PAH (APAH), which accounts for ~45% of PAH cases. TET2 mutations, a cause of clonal hematopoiesis of indeterminant potential (CHIP), predispose to an inflammatory type of PAH. We now examine mutations in another CHIP gene, DNMT3A, in PAH. Methods: We assessed DNMT3A mutation prevalence in PAH Biobank subjects as compared with controls, first using whole exome sequencing (WES)-derived CHIP calls in 1832 PAH Biobank patients versus 7509 age- and sex-matched gnomAD controls. We then performed deep, targeted panel sequencing of CHIP genes on a subset of 710 PAH Biobank patients and compared the prevalence of DNMT3A mutations therein to an independent pooled control cohort (N = 3645). In another cohort of 80 PAH patients and 41 controls, DNMT3A mRNA expression was studied in peripheral blood mononuclear cells (PBMCs). Finally, we evaluated the development of PAH in a conditional, hematopoietic, Dnmt3a knockout mouse model. Results: DNMT3A mutations were more frequent in PAH cases versus control subjects (OR 2.60, 95% CI: 1.71-4.27) when assessed in the WES dataset. Among PAH patients, 33 had DNMT3A variants, most of whom had APAH (21/33). While 21/33 had somatic mutations (female:male 17:4), germline variants occurred in 12/33 (female:male 11:1). Hemodynamics were comparable with and without DNMT3A mutations (mPAP=58+/-21 vs. 52+/-18 mmHg); however, patients with DNMT3A mutations were unresponsive to acute vasodilator testing. Targeted panel sequencing identified that 14.6% of PAH patients had CHIP mutations (104/710), with DNMT3A accounting for 49/104. There was a significant association between all CHIP mutations and PAH in analyses adjusted for age and sex (OR 1.40, 95% CI: 1.09-1.80), though DNMT3A CHIP alone was not significantly enriched (OR:1.15, 0.82-1.61). DNMT3A expression was reduced in patient-derived versus control PAH-PBMCs. Spontaneous PAH developed in Dnmt3a-/- mice and this was exacerbated by 3-weeks of hypoxia. Dnmt3a-/- mice had increased lung macrophages and elevated plasma IL-13. The IL-13; antibody canakinumab attenuated PAH in these mice. Conclusions: Germline and acquired DNMT3A variants predispose to PAH in humans. DNMT3A mRNA expression is reduced in human PAH PBMCs. Hematopoietic depletion of Dnmt3a causes inflammatory PAH in mice. DNMT3A is a novel APAH gene and may be a biomarker and therapeutic target.