Medicinal chemistry of hERG optimizations: Highlights and hang-ups.

In recent years, withdrawal of marketed drugs including sertindole, grepafloxacin, and terfenadine owing to prolongation of the length of time between the start of the Q wave and end of the T wave on an electrocardiogram (QT interval) has prompted considerable effort in trying to establish the molecular basis of this potentially lethal phenomenon. In the majority of cases, those chemical entities that prolong the QT interval and lead to torsades de pointes (TdPa) preferentially interact with a product of the human ether-a-go-go related gene (hERG), the R-subunit of IKr channels responsible for the rapid component of the delayed rectifier potassium current in the heart.1-3 Consequently, throughout the pharmaceutical industry efforts to predict QT prolongation risk have been focused on assays testing in vitro hERG channel activity in mammalian cell lines expressing the hERG channel. Medicinal chemistry groups engaged in both hit-to-lead and lead optimization activities have encountered blockade of the hERG channel as a significant hurdle along the drug discovery trajectory. Despite this, an increasing body of information has been accumulated on the strategy and tactics for overcoming inhibition of hERG. A brief summary of the in silico, in vitro, and in vivo approaches employed to measure blockade of hERG and QT prolongation is detailed in this report. The purpose of the following discussion is to summarize the approaches engaged to circumvent activity at hERG, identified through an extensive literature survey. The optimizations are recorded as pairs of compounds, which have been categorized in terms of the tactics employed to diminish hERG activity. Examination of the properties of the compounds within each category has enabled formulation of some empirical guidelines. Conclusions and recommendations for future activities are also presented. 1.1. In Vitro and In Vivo Approaches. The biophysics of cardiac action potential and relation to electrocardiogram (ECG) has been discussed extensively elsewhere.4 It is well established that blockade of the hERG channel can lead to prolongation of the action potential duration and consequently an increase in the QT interval. It should be noted, however, that the cardiac proarrhythmic toxicity of drugs through interaction with other channels or subunits involved in repolarization of heart muscles (e.g., minK, Kir2 channels) to the best of our knowledge has not yet been reported but cannot be excluded. Nevertheless, the majority of drugs that cause TdP also substantially inhibit the hERG channel at therapeutic concentrations;2 hence, significant effort has been invested in quantifying drug-hERG interaction. The 4-fold symmetry of the hERG subunit, together with the possibility of multiple binding sites, means the channel is particularly promiscuous with regard to drug interaction. This adds an extra layer of complexity in terms of the types of assay used in order to definitively assess the extent of block a drug candidate causes. Table 1 summarizes the widely applied techniques used to assess hERG blockade in both in vitro and in vivo settings. 1.2. In Silico Approaches. Several recent review articles have detailed the state of the art in terms of in silico hERG modeling.4,12-14 Approaches can be broadly classified into * To whom correspondence should be addressed. For C.J.: phone, +441698-736-190; fax, +44-1698-736-187; e-mail, [email protected]. For Z.R.: phone, +44-1698-736-157; fax, +44-1698-736-187; e-mail, [email protected]. a Abbreviations: hERG, human ether-a-go-go gene related product; TdP, torsades de pointes; minK, minimal potassium channel; KvLQT1, potassium voltage-gated channel subfamily KQT member 1; Kir2, inward rectifier potassium channel 2; IKr, potassium voltage-gated channel ether-a-go-go protein; IKs, slow voltage producing potassium channel â subunit; KcsA, Streptomyces liVidans potassium channel; MthK, Methanobacterium thermoautotrophicum potassium channel; KvAP, Aeropyrum pernix voltagedependent potassium channel; APD, action potential duration; MAP, monophasic action potential; GRIND, grid independent descriptors; MLR, multiple linear regression; DSM, discrete structural modifications; ZI, zwitterion. © Copyright 2006 by the American Chemical Society