In recent years, many drugs have been withdrawn from the market due to sudden cardiac deaths. In the pre-clinical research phase of drugs, 24% of the drugs were discontinued due to cardiovascular side effects, and 45% of the drugs were withdrawn from the market due to cardiac side effects. The main reasons for drug-induced cardiotoxicity are: blocking the fast delayed rectifier current (IKr) of the heart, causing the QT interval to be prolonged in the cardiac action potential time course, and then inducing torsion de pointes (TdP), which can be severe in severe cases. Cause sudden death. IKr is conducted by the Kv11.1 potassium ion channel encoded by the hERG gene and plays a vital role in the entire action potential time course. Drug regulatory authorities in various countries require that new chemical entities must carry out comprehensive hERG activity in accordance with the International Coordination Conference (ICH) guidelines And QT interval evaluation. Effectively predict, evaluate, and optimize as early as possible to avoid the inhibitory activity of drugs on hERG potassium channels, which will help reduce the cost of drug development and increase the success rate of drug development.
At present, solving the problem of drug hERG inhibition by optimizing the structure of lead compounds is still the most direct and effective strategy to improve cardiotoxicity. Structural optimization strategies to reduce hERG inhibition include: reducing fat solubility, reducing alkalinity, introducing hydrogen bond acceptors or groups that can generate negative ions, and conformational restrictions [1].
ACD/Percepta is a professional drug prediction software that can predict physicochemical properties, ADME properties, and toxicity properties based on the structural formula of compounds, including the prediction of cardiotoxicity—hERG inhibitory properties, which can help researchers quickly in the early stage of drug development , Accurately assess the potential cardiotoxicity risk of the compound, and guide the structural optimization of the lead. At present, ACD/Percepta has released the latest version v2020.1.0, which has greatly optimized the prediction model of hERG inhibition properties. Below, we will introduce the updated prediction module through practical examples.
Compounds 1 and 2 are a class of tertiary amine T-type calcium channel blockers reported by Shin et al. [2]. They are potential lead compounds for the treatment of cardiovascular diseases, but their hERG inhibitory activity is strong, respectively. 0.18 and 1.38 μmol·L−1, so the authors hope to modify their structure to obtain compounds with lower hERG inhibitory activity. Through different design strategies, the author synthesized six compounds (3, 4, 5, 6, 7, 8), as shown below.
Then, the author carried out the experimental determination of the activity and hERG inhibitory properties of these six compounds, and found that the hERG inhibitory effect of compounds 3 and 4 was reduced, but the activity was also significantly reduced; the hERG inhibitory effect of compounds 5 and 6 was significantly reduced , And the activity basically maintained the original level; the results of compounds 7 and 8 were disappointing. Compared with compound 2, its activity was significantly reduced and the hERG inhibitory effect was not significantly reduced, as shown in the following table.
Based on the examples in the above literature, we use Percepta software to predict the hERG inhibitory properties of the compound to see how it performs. In the latest version of Percepta V2020.1.0, on the basis of the original GALAS model, the PhysChem model is added. The algorithms of the two prediction models and the result information provided will be different, but both will give a hERG inhibitory effect (Ki < 10 μM) probability value (Probability), the value is between 0-1, the larger the value, the greater the probability of hERG inhibition (Ki <10 μM). In the GALAS model prediction results, a reliability coefficient (RI value) is also provided. The coefficient is between 0-1. The larger the value, the more reliable the prediction result.
From the above prediction results, it can be found that compounds 1 and 2 have a greater hERG inhibition risk (probability is 0.92, 0.93, respectively), and the risk of compounds 3 and 4 is relatively lower (probability is 0.80, 0.78, respectively), and compounds 5 and 6 The possibility of inhibiting hERG is significantly reduced (probability is 0.17, 0.18, respectively), while the possibility of inhibiting hERG of compounds 7 and 8 is equivalent to that of 1 and 2. Moreover, the reliability coefficients of the prediction results of this series of compounds are all high (>0.5), so the results are more credible, and in fact they are consistent with experimental results.
Physchem model
In the Physchem model interface, the software will give the prediction results of physical and chemical parameters related to hERG inhibition properties, including LogP, pKa (Acid/Base), and molecular weight (MW). Users can choose different experimental methods for simulation, including:
Conventional patch-clamp
Automated patch-clamp
Dofetilide replacement
Astemizole replacement
MK-400 replacement
Other/Not specified
In this case, we choose the Automated patch-clamp method for simulation. In addition, the prediction interface also visualizes the relationship between different physical and chemical parameters (logP, pKa) and hERG inhibition risk by means of heat maps, allowing users to more intuitively understand the risk position of the compound, which is more conducive to guiding structural optimization.
The following are the predicted results of 8 compounds:
The results show that compounds 1, 2, 7, and 8 all have a higher risk of hERG inhibition (Ki <10 μM), and compounds 3, 4, 5, and 6 have relatively low risks, which can be seen more intuitively in the heat map. Figure out the risk position of each compound. In addition, the Physchem model allows users to input experimental data of physical and chemical parameters such as logP, pKa, etc., so that the software can perform simulation prediction again, and obtain more accurate prediction results of hERG inhibition properties.
In short, the Physchem model introduces empirical rules for the main physical and chemical parameters that affect hERG inhibition (such as reducing fat solubility, reducing alkalinity, introducing hydrogen bond receptors or groups that can generate negative ions can reduce hERG inhibition) to make it quantitative, Visualization is more helpful for the optimization guidance of the leader!
The above is the test of compounds reported in the literature through the hERG inhibitory properties prediction function of Percepta software. The results show that the prediction results of the two hERG inhibitory properties prediction models (GALAS model and Physchem model) of Percepta software are basically consistent with the experimental results. In the drug design stage, the use of software prediction methods for compound toxicity risk assessment can avoid unnecessary failures (such as compounds 7 and 8 in the literature). Of course, we can also use the Structrue Designer function of the Percepta software to design more derivatives, explore a larger chemical space, and find candidate compounds with better druggability, starting from the lead, and then perform synthesis, bioassay, and medicine. Generation experiments or safety tests, thereby reducing the failure rate of drug development.
references:
1. Lead compound structure optimization strategy (5)-Reduce the drug hERG cardiotoxicity. Acta Pharmaceutica Sinica 2016, 51 (10): 1530 −1539.
2. Successful reduction of off-target hERG toxicity by structural modification of a T-type calcium channel blocker. Bioorganic & Medicinal Chemistry Letters 24 (2014) 880–883.
hERG test-preclinical cardiac safety evaluation
hERG Assay – Drug Preclinical Cardiac Safety Assessment
Share：
marketing@medicilon.com
