Introduction: Atrial fibrillation (AF) frequently occurs alongside heart failure (HF), however, their interaction remains poorly understood. Clinical studies involving patients with AF and HF revealed improved left ventricular (LV) function and reduced mortality following AF-catheter ablation. However, the underlying mechanisms behind the improvements following rhythm restoration remain unclear. Previous studies from our group suggested the involvement of reactive oxygen species (ROS) in AF-induced alterations of cardiomyocyte Ca2+ handling processes, but recovery has not been investigated yet.

 

Objective: This study investigated effects of AF and its termination on human LV myocardium as well as potential molecular mechanisms involved in the AF-induced electrical remodeling.

Methods and Results: Twitching slices of human LV tissue from end-stage HF patients were long-term cultivated and examined in vitro. AF simulation was performed using tachy-arrhythmic culture pacing at 100 bpm with 30% beat-to-beat variability (compared to sinus rhythm simulation with 60 bpm/0% variability) resulting in a progressive decrease in systolic force in human LV myocardium (n=9, 5 hearts) with a significantly reduced twitch amplitude after 7 days of AF simulation compared to the control (n=6/5). Diastolic tension remained unchanged. Rhythm restoration was simulated by switching from AF to sinus rhythm simulation. (n=6/4) after 7 days. Remarkably, within just 5 days after AF termination, a significant improvement in LV function was observed, indicated by recovered systolic contraction amplitudes. 

To elucidate potential mechanisms in human cardiomyocytes, we utilized human iPSC-cardiomyocytes (iPSC-CM) (n=4 differentiations). Epifluorescence microscopy measurements using Fura-2 staining revealed a significant reduction in systolic Ca2+ transient amplitudes after 7 days of AF simulation (100bpm/30%, n=41) compared to the control (60 bpm/0%, n=44). Notably, diastolic Ca2+ levels remained unaffected. However, after 5 days of sinus rhythm following AF simulation, iPSC-CM exhibited no difference in Ca2+ transient amplitude (n=41) anymore compared to the control group (n=28), indicating a fast recovery of systolic Ca2+ cycling. As an underlying mechanism of systolic Ca2+ transient regulation we identified an increased diastolic Ca2+ release after 7 days of AF simulation (n=19) compared to control (n=27) in confocal microscopy line-scans (Fluo-4), which was subsequently reversed after 5 days of sinus rhythm in iPSC-CM (n=15, control n=16). These mechanisms could potentially explain the contractile response to AF in human myocardium. 

To investigate the impact of reactive ROS on the AF-induced electrical remodeling of the LV, we subjected iPSC-CM to treatment with the ROS-scavenger N-acetylcysteine (NAC). Cells treated with NAC showed significantly increased Ca2+ transient amplitudes in both experimental groups (Ctrl n=43, AF n= 35) as well as in recovered cells (Ctrl n=27, AF n= 35) compared to the control (Ctrl n=74, AF n=89), suggesting an interplay of ROS with the Ca2+ homeostasis of cardiomyocytes.

Conclusion: This study demonstrates that AF per se impairs human LV function in HF patients. The involved changes in cardiomyocyte Ca2+ handling are reversible upon AF termination leading to improved LV function already within few days after AF cessation. Identification of these mechanisms may help to treat contractile dysfunction in patients with HF and permanent AF.