Distinct contractile and molecular differences between two goat models of atrial dysfunction: AV block-induced atrial dilatation and atrial fibrillation.

Atrial dilatation is an independent risk factor for thromboembolism in patients with and without atrial fibrillation (AF). In many patients, atrial dilatation goes along with depressed contractile function of the dilated atria. While some mechanisms causing atrial contractile dysfunction in fibrillating atria have been addressed previously, the cellular and molecular mechanisms of atrial contractile remodeling in dilated atria are unknown. This study characterized in vivo atrial contractile function in a goat model of atrial dilatation and compared it to a goat model of AF. Differences in the underlying mechanisms were elucidated by studying contractile function, electrophysiology and sarcoplasmic reticulum (SR) Ca2+ load in atrial muscle bundles and by analyzing expression and phosphorylation levels of key Ca2+-handling proteins, myofilaments and the expression and activity of their upstream regulators. In 7 chronically instrumented, awake goats atrial contractile dysfunction was monitored during 3 weeks of progressive atrial dilatation after AV-node ablation (AV block goats (AVB)). In open chest experiments atrial work index (AWI) and refractoriness were measured (10 goats with AVB, 5 goats with ten days of AF induced by repetitive atrial burst pacing (AF), 10 controls). Isometric force of contraction (FC), transmembrane action potentials (APs) and rapid cooling contractures (RCC, a measure of SR Ca2+ load) were studied in right atrial muscle bundles. Total and phosphorylated Ca2+-handling and myofilament protein levels were quantified by Western blot. In AVB goats, atrial size increased by 18% (from 26.6+/-4.4 to 31.6+/-5.5 mm, n=7 p<0.01) while atrial fractional shortening (AFS) decreased (from 18.4+/-1.7 to 12.8+/-4.0% at 400 ms, n=7, p<0.01). In open chest experiments, AWI was reduced in AVB and in AF goats compared to controls (at 400 ms: 8.4+/-0.9, n=7, and 3.2+/-1.8, n=5, vs 18.9+/-5.3 mmxmmHg, n=7, respectively, p<0.05 vs control). FC of isolated right atrial muscle bundles was reduced in AVB (n=8) and in AF (n=5) goats compared to controls (n=9) (at 2 Hz: 2.3+/-0.5 and 0.7+/-0.2 vs 5.5+/-1.0 mN/mm2, respectively, p<0.05). APs were shorter in AF, but unchanged in AVB goats. RCCs were reduced in AVB and AF versus control (AVB, 3.4+/-0.5 and AF, 4.1+/-1.4 vs 12.2+/-3.2 mN/mm2, p<0.05). Protein levels of protein kinase A (PKA) phosphorylated phospholamban (PLB) were reduced in AVB (n=8) and AF (n=8) vs control (n=7) by 37.9+/-12.4% and 29.7+/-10.1%, respectively (p<0.01), whereas calmodulin-dependent protein kinase II (CaMKII) phosphorylated ryanodine channels (RyR2) were increased by 166+/-55% in AVB (n=8) and by 146+/-56% in AF (n=8) goats (p<0.01). PKA-phosphorylated myosin-binding protein-C and troponin-I were reduced exclusively in AVB goat atria (by 75+/-10% and 55+/-15%, respectively, n=8, p<0.05). Atrial dilatation developing during slow ventricular rhythm after complete AV block as well as AF-induced remodeling are associated with atrial contractile dysfunction. Both AVB and AF goat atria show decreased SR Ca2+ load, likely caused by PLB dephosphorylation and RYR2 hyperphosphorylation. While shorter APs further compromise contractility in AF goat atria, reduced myofilament phosphorylation may impair contractility in AVB goat atria. Thus, atrial hypocontractility appears to have distinct molecular contributors in different types of atrial remodeling.