Ionic Currents and Action Potential Prolongation
IKr is generated by expression of the human ether a-go-go-related gene (HERG, now termed KCNH2). The electrophysiological characteristics of heterolo-gously expressed KCNH2 are very similar, but not identical to, IKr recorded from human cells; one commonly observed difference is in the rates of deactivation. A commonly invoked explanation for this discrepancy is that KCNH2 associates with other protein(s) in at least some (but perhaps not all) human myocytes to generate IKr. Candidate function-modifying proteins include members of the KCNE family (notably KCNE2) (Abbott et al. 1999), as well as the a-subunit that generates IKs (KCNQ1) (Ehrlich et al. 2004). Differences in post-translational modification may represent another mechanism. Site-directed mutagenesis and structural modeling studies have identified key features of the HERG/KCNH2 protein, absent in other potassium channels,
that seem to underlie the fact that many structurally unrelated drugs inhibit IKr (Mitcheson et al. 2000). Drugs block the channel by accessing the pore region from the intracellular side of the channel. The HERG/KCNH2 protein, unlike other K+ channels, lacks proline groups within S6, and the resultant lack of "kinking" of the S6 region is thought to facilitate access of even relatively bulky drugs to the pore region. In addition, the S6 also includes two aromatic residues, absent in other K+ channels, oriented to face the pore, and these are thought to provide high-affinity drug binding sites with many drugs. Since the channel is a tetramer, there are actually eight such potential high-affinity sites within the pore, a feature also absent in other K+ channels.
Screening new drugs for IKr block can be done using myocytes from a number of mammalian species (dog, rabbit, cat, guinea pig; but not adult mouse or rat), in cultured neonatal mouse cells (ATI or HL1 cells), or by heterologous expression of HERG/KCNH2. The results obtained in such studies can generally define whether or not a drug is a potent blocker of the current, one important component of assessing the balance of potential risk versus anticipated benefit for a new drug.
When conditions mimicking those seen in torsades de pointes (hypokalemia, slow drive rates, QT-prolonging drug) are used in vitro, action potentials in cells of the conduction system (Purkinje fibers) markedly prolong and generate distinctive discontinuities and spontaneous upstrokes, arising from phase 3 of the action potential (Strauss et al. 1970; Dangman and Hoffman 1981; Roden and Hoffman 1985). These events are termed early afterdepolarizations (EADs), distinguishing them from DADs that arise after the action potential is fully repolarized. The ionic current that underlies the upstroke represented by EADs has not been fully defined. In some experiments, it is clear that reactivation of L-type calcium channels (enabled by the long phase 2 of prolonged action potentials) contributes (January and Riddle 1989). Indeed, in in vitro experiments L-type calcium channel blockers are highly effective in eliminating the triggered upstroke and reducing action potential prolongation (Nattel and Quantz 1988). Nevertheless, these agents have not been terribly effective at preventing torsades de pointes (although there is no randomized prospective trial). Other evidence points to a role for intercellular calcium overload and an Iti-like mechanism, especially for EADs arising during phase 3 of the action potential (Wu et al. 1999). For example, although EADs are generally considered to be "bradycardia-dependent," they can also be elicited by rate acceleration, followed by a brief pause (Burashnikov and Antzelevitch 1998).
Post a comment