Cardiotoxicity Studies in Zebrafish - Zebrafish: Methods for Assessing Drug Safety and Toxicity - page 59

We were able to confirm heightened sensitivity of heterozygote breakdance
mutants to I Kr blockade that would be predicted from current understanding of
repolarization reserve (Roden, 2004). At baseline, breakdance heterozygotes display
minimal action potential prolongation, but exposure to the potent and specific I Kr
blocker dofetilide at 10 nM (similar to circulating plasma levels of the drug in clinical
use) prolonged their action potentials (Fig. 5.3c). In order to establish that other
mechanisms of reploarization toxicity are accessible in the zebrafish model, we also
exposed zebrafish to sea anemone toxin (ATX-II), a toxin that interferes with sodium
channel inactivation (Catterall and Beress, 1978). This resulted in a dose-dependent
increase in the action potential duration, extending the parallels with human QT
interval to include the LQT3 syndrome (Fig. 5.3d).
Subsequent work using optical mapping has also confirmed that novel repolar-
ization genes such as NOS1AP can be efficiently studied in the zebrafish. These data
suggest that ongoing work by us, and others, to define novel predictive algorithms for
cardiotoxicity is likely to be useful in preclinical testing. Recently, we have obtained
evidence that pharmacogenetic discovery in the zebrafish may predict human
repolarization genes. In a shelf screen using staged assays (heart rate, atrioventricular
block, and optical mapping) of insertional mutants, we implicated 15 novel genes in
cardiac repolarization, one of which has recently been identified as a modifier of the
human QT interval (Milan et al., 2009).
5.5 FUTURE DIRECTIONS
By building assays that exploit the throughput and the fidelity of the zebrafish, it
should be feasible to define a truly predictive preclinical toxicology that encompasses
not only complex effects on native cardiac biology, but also absorption, distribution
metabolism, and excretion. Clearly, it will be vital to establish robust correlations
between each phenotype in larval zebrafish and the cognate clinical trait, but it is
evident from work in repolarization to date that the fish may be at least as
representative as traditional models. The throughput feasible in multiwell plates
opens up in vivo screening to a scale of investigation previously only accessible in
vitro. Preliminary observations also suggest that gender effects observed in human
drug-induced cardiotoxicity may be accessible (Pham and Rosen, 2002). Possible
differences in the mechanisms or scope of action of different “I Kr blockers” and drug
interaction mechanisms can be addressed with this whole animal model. The
development of sensitized or resistant zebrafish reporter strains will improve the
specificity of this system. Finally, human pharmacogenetic studies of cardiac
repolarization have been limited to the evaluation of candidate genes, as more
powerful segregation-based family studies of drug responses are not feasible (Sesti
et al., 2000; Roden and George, 2002; Splawski et al., 2002; Yang et al., 2002). The
ease of genetic manipulation in the zebrafish should allow the unbiased identification
of inherited modifiers of drug responses in phenotype-driven screens. By combining
assays of increasing resolution in a tiered approach it is possible to optimize
throughput and biological representation, and thus allow predictive toxicology to