Abstract-Although the canine atrium has proven useful in several experimental models of atrial fibrillation and for studying the effects of rapid atrial pacing on atrial electrical remodeling, it may not fully represent the human condition because of reported differences in functional ionic currents and ion channel subunit expression. In this study, we reassessed the molecular components underlying one current, the ultrarapid delayed rectifier current in canine atrium [I Kur(d) ], by evaluating the mRNA, protein, immunofluorescence, and currents of the candidate channels. Using reverse transcriptase-polymerase chain reaction, we found that Kv1.5 mRNA was expressed in canine atrium whereas message for Kv3.1 was not detected. Western analysis on cytosolic and membrane fractions of canine tissues, using selective antibodies, showed that Kv3.1 was only detectable in the brain preparations, whereas Kv1.5 was expressed at high levels in both atrial and ventricular membrane fractions. Confocal imaging performed on isolated canine atrial myocytes clearly demonstrated the presence of Kv1.5 immunostaining, whereas that of Kv3.1 was equivocal. Voltage-and current-clamp studies showed that 0.5 mmol/L tetraethylammonium had variable effects on sustained K ϩ currents, whereas a compound with demonstrated selectivity for hKv1.5 versus Kv3.1, hERG or the sodium channel, fully suppressed canine atrial I Kur tail currents and depressed sustained outward K ϩ current. This agent also increased action potential plateau potentials and action potential duration at 20% and 50% repolarization. These results suggest that in canine atria, as in other species including human, Kv1.5 protein is highly expressed and contributes to I Kur .

Na(+) conductance through cloned K(+) channels has previously allowed characterization of inactivation and K(+) binding within the pore, and here we have used Na(+) permeation to study recovery from C-type inactivation in human Kv1.5 channels. Replacing K(+) in the solutions with Na(+) allows complete Kv1.5 inactivation and alters the recovery. The inactivated state is nonconducting for K(+) but has a Na(+) conductance of 13% of the open state. During recovery, inactivated channels progress to a higher Na(+) conductance state (R) in a voltage-dependent manner before deactivating to closed-inactivated states. Channels finally recover from inactivation in the closed configuration. In the R state channels can be reactivated and exhibit supernormal Na(+) currents with a slow biexponential inactivation. Results suggest two pathways for entry to the inactivated state and a pore conformation, perhaps with a higher Na(+) affinity than the open state. The rate of recovery from inactivation is modulated by Na(+)(o) such that 135 mM Na(+)(o) promotes the recovery to normal closed, rather than closed-inactivated states. A kinetic model of recovery that assumes a highly Na(+)-permeable state and deactivation to closed-inactivated and normal closed states at negative voltages can account for the results. Thus these data offer insight into how Kv1. 5 channels recover their resting conformation after inactivation and how ionic conditions can modify recovery rates and pathways.

RSD1235 itself has minor actions on repolarization in Purkinje fibers, but can reverse the AP-prolonging actions of class III agents and terminate arrhythmias in a model of TdP. We suggest that these protective actions of RSD1235 may result, at least in part, from its ability to inhibit late I(Na) during action potential repolarization.