Cardiotoxicity Studies in Zebrafish - Zebrafish: Methods for Assessing Drug Safety and Toxicity - page 54

heterogeneity. Thus, toxicities that depend on drug interactions with specific genetic
variations may never be observed prior to human trials, and systematic studies of
gene-drug interactions are not currently performed. These limitations of conven-
tional preclinical testing have led to efforts to develop in vivo models capable of
high-fidelity and high-throughput studies of vertebrate drug response for predictive
toxicology and toxicogenetics. Many groups have focused on the zebrafish, and in
this chapter we will discuss the utility of the organism for such studies using
repolarization toxicity as an example.
5.2 REPOLARIZATION TOXICITY
Repolarization is extremely complex, depending on the integrated effects of
individual channels, receptors, cytoskeletal elements, and the membrane. There
is also extensive regional electrical heterogeneity within the heart that has been
shown to be an important contributor to arrhythmogenesis (Antzelevitch and
Fish, 2001). Further, drugs that may be safe in isolation can perturb repolarization
when given with other medications (Camm et al., 2000), through pharmacokinetic
or pharmacodynamic drug-drug interactions. Finally, genetic variation has been
shown to contribute to individual susceptibility to drug-induced arrhythmias, and
may be much more common than has been appreciated (Roden, 2001). A tractable
model system that enabled the systematic identification of genes responsible for
such variation would offer many advantages.
Virtually all of the drugs causing repolarization toxicity have been shown to
inhibit the rapid component of the repolarizing potassium current (I Kr ) in vitro. This
current is conducted by a channel composed of multiple subunits, including the
proteins KCNH2 and KCNE2. Traditional in vitro assays of I Kr are limited by the
heterologous nature of existing systems, the absence of many channel components
and interacting proteins that are important in differentiated cardiomyocytes, low
throughput, and the inability to detect drug-drug interactions that depend on other
organs (Eckardt et al., 1998; Yang et al., 2001). Animal models, while more
physiologic, have an even lower throughput, restricting their ability to screen
systematically at scale for drug-drug interactions which may involve a large
proportion of the current formulary.
The zebrafish heart exhibits both a complex repertoire of ion channels and
functioning metabolism within only 24 h of fertilization (Baker et al., 1997). The
embryo's optical properties facilitate the evaluation of heart rate and rhythm. In
addition, profoundly abnormal cardiac function is tolerated by larval fish that may
survive for 4-5 days without a circulation (Warren and Fishman, 1998).
5.3 INITIAL SCREENING: BRADYCARDIA
In initial work to evaluate 100 small molecules for their effects on embryonic
physiology, we observed that compounds known to cause QT prolongation and