Background:

Atrial arrhythmias frequently occur in patients with obstructive sleep apnea (OSA), but the underlying mechanisms remain insufficiently understood. We recently demonstrated that CaMKII-dependent pro-arrhythmic activity is increased in patients with OSA, but the mechanisms of CaMKII activation remain unknown. Interestingly, OSA can lead to increased reactive oxygen species production, which may facilitate CaMKII activation by oxidation at methionine 281/282.

Purpose:

We tested if oxidation of CaMKII is involved in the development of arrhythmias in a novel mouse model of obstructive sleep apnea by tongue enlargement, which our group has previously established.

Methods and Results:

Polytetrafluorethylene (PTFE, 100 µl) was injected into the tongue of mice to induce sustained enlargement. This intervention results in the development of obstructive sleep apnea with spontaneous inspiratory flow limitations (IFLs) as monitored by whole-body plethysmography. After eight weeks, isolated atrial and ventricular cardiomyocytes were incubated with the Ca-sensitive dye FURA-2 AM (5 µM, 15 min) and analyzed with epifluorescence microscopy. Ca transients were elicited by regular electrical field stimulation (1 Hz and 4Hz, 20 V for 4 ms). Deviations from diastolic Ca baseline between two stimulated transients were defined as pro-arrhythmic non-stimulated events (fig 1A, red arrows). Compared to untreated control (CTRL) mice, PTFE injection resulted in a significant increase in the frequency of atrial pro-arrhythmic events (s-1) at 1 Hz (from 0.02 ± 0.01 to 0.05 ± 0.01, p<0.001, N= 6 vs. 6 fig 1B). Intriguingly, genetically modified mice that lack oxidative activation of CaMKII by mutation of methionine 281/282 to valin (MMVV) were protected from arrhythmias after PTFE-treatment. Compared to WT OSA mice, the arrhythmia frequency was significantly reduced in MMVV OSA mice to 0.02 ± 0.01 (p<0.001, N= 8, fig. 1B). Preliminary analysis revealed a non-significant trend towards a positive correlation of pro-arrhythmic events with IFLs in WT mice (p=0.18), which was not present in MMVV mice. At 4 Hz, which is closer to the physiologic murine heart rate, similar effects were observed, with an increased frequency of atrial pro-arrhythmic events of 0.14 ± 0.07 in WT OSA mice, which was significantly reduced in MMVV OSA mice (p=0.004, N= 6 vs. 8). In accordance to the results in atrial myocytes, pro-arrhythmic events were also increased in ventricular myocytes of WT OSA mice to 0.07 ± 0.02 (compared to 0.04 ± 0.02 in WT CTRL mice, 1Hz, N= 4 vs. 5, P=0.008, fig 1C), and significantly reduced in ventricular myocytes of MMVV OSA mice (p<0.001, N=7, Fig.1C). Similar effects were observed at 4Hz stimulation.

Conclusion:

In a mouse model of obstructive sleep apnea, atrial and ventricular pro-arrhythmic activity is modulated by oxidative CaMKII activation, which may have therapeutic implications.