Abstract Body: Introduction: Atrial fibrillation (AF) is a leading cause of stroke and heart failure. Recent genome-wide association studies have implicated ACVR2A, the gene encoding the Activin type IIA receptor (ActRIIA) in AF pathogenesis. The role of ActRIIA signaling in AF and its potential as a therapeutic target are unknown.
Hypothesis: ActRIIA signaling is causal in AF pathophysiology.
Methods: Causality was assessed using gain- and loss-of-function approaches in mice and cardiomyocytes (CM). ActRII signaling was increased in C57BL/6 mice using an activinA adenovirus (Ad.ActA) or recombinant ActA protein (R.ActA). ActRII was inhibited with CM-specific ActRIIA or ActRIIB knockout (KO) mice or an ActRIIA-Fc ligand trap in three murine AF models induced by high-fat diet (HFD), transverse aortic constriction (TAC), or aging. AF inducibility and electrical remodeling were assessed by invasive electrophysiology and optical mapping. Cardiac RNAseq was used to identify mechanisms, which were validated in CM and mice.
Results: Ad.ActA and R.ActA increased circulating ActA, cardiac ActRII signaling, and AF inducibility (0%[Ad.GFP] vs 75%[Ad.ActA], p=0.06, n=4; 33%[PBS] vs 100%[R.ActA], p=0.001, n=9), and decreased atrial conduction velocity (0.6±0.02[Ad.GFP] vs 0.4±0.05m/s[Ad.ActA], p=0.001, n=3). Conversely, decreased trends of AF inducibility occurred in CM-ActRIIA-KO mice subjected to HFD (71%control] vs 48%[KO], p=0.09, n=21) or TAC (69%[control] vs 33%[KO], p=0.08, n=12), but not in CM-ActRIIB-KO mice. ActRIIA-Fc decreased AF inducibility in 22-month-old mice (85%[PBS] vs 0%[ActRIIA-Fc], p=0.01, n=5). RNAseq identified gap junctions as one of the most downregulated pathways in Ad.ActA mice (NES= -2.8, FDR< 0.001). In both mice and CM, R.ActA decreased CX43 protein expression by ~25% (p<0.05), which was associated with increased expression of the E3 ubiquitin ligase Smurf1 (mice: 1.3 fold, p=0.09; CM: 1.9 fold, p=0.02). In CM, ActRIIA, but not ActRIIB, knockdown (KD) decreased Smurf1 (-3-fold, p<0.001) and increased CX43 (1.8-fold, p=0.02). Lastly, both genetic and pharmacologic (A01) inhibition of Smurf1 increased CX43 in CM by 1.2-fold (p=0.06[KD], p=0.04[A01]).
Conclusion: Increased ActA/ActRII signaling is causal in AF pathogenesis. This effect is mediated through ActRIIA, but not ActRIIB, suggesting ActRIIA specificity in AF, and its potential as a therapeutic target.