The vagus slows the heart by raising potassium permeability and the activated K+ current – IK(ACh) – is carried by heteromers of the G-protein-gated inward rectifier potassium channel subunits Kir3.1 and Kir3.4 (for review, see Stanfield et al. 2002). However, in atrial myocytes only about half of Kir3.4 can be purified in heteromers with Kir3.1 and the remainder forms Kir3.4 homomers (Corey & Clapham, 1998). Quite what these homomeric channels do physiologically has been a puzzle: mice with Kir3.1 knocked out show mild tachycardia, which fails to respond to carbachol suggesting that Kir3.4 homomers do not contribute significantly to IK(ACh) (Bettahi et al. 2002). Clinical interest in the molecular biology of these channels is heightened by the need to manage atrial fibrillation, a common complication of heart failure and valvular disease. The dysrhythmia leads to fast, ‘irregularly irregular’ pacing of the ventricles, compromising cardiac output. In addition, blood is held in atria, and thrombus formation, particularly in the left auricle, increases risk of stroke. Atrial fibrillation of congestive heart failure (CHF) has often been managed with digoxin in the UK, and the drug has been held, on the basis of ECG monitoring, to have an effect via the vagus nerve, activating IK(ACh) (Krum et al. 1995). Atrial fibrillation (AF) is followed by electrophysiological remodelling of the heart, a process that is reversed after restoration of sinus rhythm (Brundel et al. 2001; Gaborit et al. 2005). In atrial myocytes, Kir3.1 expression is reduced relative to that of Kir3.4. Kir3.0 currents become constitutively active and the response to carbachol is reduced. The shortening of atrial action potentials may be corrected by the bee venom tertiapin, to which there is increased sensitivity (Dobrev et al. 2005), and use of derivatives of this venom has been seen as a possible method of management, mirroring use of the K+ channel blocker amiodarone. In this issue of The Journal of Physiology, Mintert et al. (2007) address some of these issues, using over-expression of Kir3.4 to mimic part of what happens in persistent AF. The homomeric channels produced show weaker inward rectification than is seen with Kir3.1/3.4 heteromers (the residues binding gating polyamines come from both Kir3.1 and 3.4, e.g. Stanfield et al. 2002). Unlike Kir3.1/3.4 heteromers, which must be gated open by Gβγ, these homomers are constitutively active. Nonetheless, FRET studies show homomers do interact with Gβγ. Kir3.0 channels require the membrane lipid phosphatidyl inositol bis phosphate (PIP2) as cofactor. Channels are thought to be activated by intracellular Na+ (Sui et al. 1998), Na+ binding to an aspartate residue in the C-terminus (Ho & Murrell-Lagnado, 1999) of Kir3.4. (The residue is Asn in Kir3.1.) By masking the negative charge of aspartate, Na+ aids the binding of PIP2. Surprisingly in view of this, the paper by Mintert et al. (2007) shows that IK(ACh) of atrial myocytes, carried by Kir3.1/3.4 heteromers, is not activated by internal Na+ at concentrations up to 60 mm; but Kir3.4 homomers are highly sensitive to Na+. The sensitivity to Na+ does not require activation through Gβγ since GTP-β-S has little inhibitory effect. Na+-activated currents are little inhibited by partial depletion of PIP2: a higher affinity probably means that PIP2 can bind from a depleted pool. Finally, channels are exquisitely sensitive to tertiapin-Q, affinity being at least 10-fold greater than that of Kir3.1/Kir3.4 heteromers. Do these findings concerning activation only of homomeric channels, which come to dominate over heteromers in atrial fibrillation, explain how digoxin controls AF? The drug might raise internal Na+ and activate Kir3.4 currents. In fact Mintert et al. (2007) are careful not to draw this conclusion, which would be controversial in the light of clinical literature describing the shortening of atrial action potentials as central to AF, with derivatives of tertiapin being seen as a solution to management. Further studies of this common dysrhythmia are sure to follow.
Read full abstract