Predictive models for hERG potassium channel blockers

doi:10.1016/j.bmcl.2005.03.062

Bioorganic & Medicinal Chemistry Letters

Volume 15, Issue 15, 1 August 2005, Pages 3637-3642

https://doi.org/10.1016/j.bmcl.2005.03.062 Get rights and content

Abstract

We report here a general method for the prediction of hERG potassium channel blockers using computational models generated from correlation analyses of a large dataset and pharmacophore-based GRIND descriptors. These 3D-QSAR models are compared favorably with other traditional and chemometric based HQSAR methods.

Graphical abstract

Computational QSAR models constructed from pharmacophore-based GRIND descriptors were found to be predictive for hERG channel blockers.

Section snippets

Additional material

Compound	Experimental (pIC₅₀–6)	Calculated (pIC₅₀–6)
Sertindole	2.52	2.11
mol10^a	2.21	1.87
mol3^a	2.15	1.74
Dofetilide	2.00	1.94
mol2^a	2.00	1.62
mol15^a	1.96	0.33
Pimozide^a	1.77	1.42
Sparfloxacin	1.74	1.56
mol13^a	1.63	1.10
Haloperidole	1.49	0.86
MK499	1.47	1.61
mol8^a	1.44	1.35
Cisapride	1.35	1.01
Terfenadine	1.25	0.83
mol1^a	1.06	1.42
mol7^a	0.88	1.09
mol6^a	0.86	0.66
Verapamil	0.84	0.89
Ziprasidone	0.77	0.88
Thioridazine	0.72	0.91
Halofantrine^b	0.71	0.54
mol14^a	0.69	0.59
mol11^a	0.34	0.18
Quinidine	0.26	0.20
Azimilide^b	0.25	0.15
mol4	0.24	0.04
Cyamemazine	0.06	0.01
Chlorpromazine^b

Acknowledgments

We wish to thank Stephan Reiling for help with database manipulation.

References and notes (28)

M.C. Sanguinetti et al.
Cell
(1995)
W. Tang et al.
J. Biomol. Screen.
(2001)
K. Finlayson et al.
Eur. J. Pharmacol.
(2001)
P.J. Chiu et al.
J. Pharmacol. Sci.
(2004)
J. Kang et al.
Eur. J. Pharmacol.
(2000)
R.A. Pearlstein et al.
Bioorg. Med. Chem. Lett.
(2003)
A.M. Aronov et al.
Bioorg. Med. Chem.
(2004)
G.M. Keseru
Bioorg. Med. Chem. Lett.
(2003)
Y.V. Korolkova et al.
J. Biol. Chem.
(2002)
L. Pardo-Lopez et al.
J. Biol. Chem.
(2002)

F. De Ponti et al.

Drug Safety

(2002)

J.S. Mitcheson et al.

Curr. Opin. Drug Discov. Dev.

(2003)

R. Pearlstein et al.

J. Med. Chem.

(2003)

R. Netzer et al.

Curr. Opin. Drug Discov. Dev.

(2003)

Cited by (96)

Multi-algorithm based machine learning and structural pattern studies for hERG ion channel blockers mediated cardiotoxicity prediction
2021, Chemometrics and Intelligent Laboratory Systems
Citation Excerpt :
Earlier research showed that numerous in silico based approaches such as classification, QSAR, pharmacophore mapping, virtual screening, homology modelling, etc. have been published on hERG ion channel blockers modelling and its structural analysis. The quantitative methods provide useful information on the structure activity relationship (SAR) and the classification methods used to segregate the compounds as actives or inactives, still both these techniques are certainly useful [5,11–19]. However, the construction and utilization of significant in silico approaches for the hERG ion channel blocker modelling is still in hunting.
The development of novel bioactive molecules without hERG ion channel blocking activities is one of the important tasks for the drug discovery scientists. Both experimental and in silico techniques are applied for screening/assaying of novel molecules and existing molecules against hERG ion channel to study its cardiotoxicity. Hence, in the present investigation, QSAR and multialgorithm based classification analysis such as Random Forest (RF), Decision Tree (DT), Support Vector Machine (SVM) and Naive Bayesian (BN) were performed on the data set possessed hERG ion channel blocking activity using different variable (MOE-descriptors and MACCS Fingerprints). The models developed with the MLR and the machine learning techniques provided significant results which were confirmed by their acceptable statistical parameter values. Further, the sum of ranking differences (SRDs) calculated for the models showed that the Model 6-MOE, Model 7-MOE, Model 4-SVM, Model 7-SVM and Model 9-SVM are having smaller SRD values and are placed as top ranked models. The MOE-descriptors and the MACCS Fingerprints contributed in these models revealed that the volume-surface area properties, negative charge on the vdW surface area, number of fluorine atom, number of aromatic ring/atoms, less number of oxygen and hydrogen bonding and donor atoms are contributed for the activity. Furthermore the MACCS Fingerprints such as MACCSFP42, MACCSFP85, MACCSFP122 and MACCSFP139 also explain the polarity of the compounds. The structural pattern analysis with the MACCS Fingerprint revealed that the aromatic rings responsible for hydrophobicity, halogen and heteroatoms provide polarity to the compounds and the asymmetric carbon make the compounds are highly flexible. These studies concluded that the derived results can be used for prediction of hERG ion channel blocking activity of novel molecules.
Computational prediction of hERG blockers using homology modelling, molecular docking and QuaSAR studies
2021, Results in Chemistry
A full-length three-dimensional structure of the tetrameric potassium ion channel (hERG Kv11.1) including the N- and C-terminal domains was built, with a diameter of 6 Å and 12 Å between the K + selectivity filter and the pore cavity residue Tyr⁶⁵² of opposite subunits. Further docking studies with a set of 233 structurally known blockers have shown that compounds bind near the inner vestibule of the pore channel, as well as the helix-IV region of the voltage sensor domain (VSD) in the alpha subunit. The residues of hERG, Gly⁶²⁶, Phe⁶²⁷, Gly⁶²⁸, Tyr⁶⁵² and Phe⁶⁵⁶ of the pore channel and Arg⁴⁸⁸ of VSD plays an important role in ligand binding and hERG blockage. The conducted QuaSAR model is statistically significant, with R² of 0.72 in predicting the hERG blocking activity. Furthermore, QuaSAR descriptors employing computer-assisted multiple regression procedure reveal that increase in hydrophobicity with higher number of aromatic rings are favorable for the binding affinity of hERG blockers. Additionally, the pIC₅₀ values of 25 commercial compounds screened using structure-based pharmacophore model also show binding to the selectivity filter and pore cavity of hERG potassium channel like the known hERG blockers with a wide range of inhibition from weak to strong blockage predicting to have proarrhythmic potential.
Performance of machine learning algorithms for qualitative and quantitative prediction drug blockade of hERG1 channel
2018, Computational Toxicology
Drug-induced abnormal heart rhythm known as Torsades de Pointes (TdP) is a potential lethal ventricular tachycardia found in many patients. Even newly released anti-arrhythmic drugs, like ivabradine with HCN channel as a primary target, block the hERG potassium current in overlapping concentration interval. Promiscuous drug block to hERG channel may potentially lead to perturbation of the action potential duration (APD) and TdP, especially when combined with polypharmacy and/or electrolyte disturbances. The example of novel anti-arrhythmic ivabradine illustrates clinically important and ongoing deficits in drug design and warrants for better screening methods. There is an urgent need to develop new approaches for rapid and accurate assessment of how drugs with complex interactions and multiple subcellular targets can predispose or protect from drug-induced TdP. One of the unexpected outcomes of compulsory hERG screening implemented in USA and European Union resulted in large datasets of IC₅₀ values for various molecules entering the market. The abundant data allows now to construct predictive machine-learning (ML) models. Novel ML algorithms and techniques promise better accuracy in determining IC₅₀ values of hERG blockade that is comparable or surpassing that of the earlier QSAR or molecular modeling techniques. To test the performance of modern ML techniques, we have developed a computational platform integrating various workflows for quantitative structure activity relationship (QSAR) models using data from the ChEMBL database. To establish predictive powers of ML-based algorithms we computed IC₅₀ values for large dataset of molecules and compared it to automated patch clamp systems for a large dataset of hERG blocking and non-blocking drugs, an industry gold standard in studies of cardiotoxicity. The optimal protocol with high sensitivity and predictive power is based on the novel eXtreme gradient boosting (XGBoost) algorithm. The ML-platform with XGBoost displays excellent performance with a coefficient of determination of up to R² ∼0.8 for pIC₅₀ values in evaluation datasets, surpassing other metrics and approaches available in literature. Ultimately, the ML-based platform developed in our work is a scalable framework with automation potential to interact with other developing technologies in the cardiotoxicity field, including high-throughput electrophysiology measurements delivering large datasets of profiled drugs, rapid synthesis and drug development via progress in synthetic biology.
Computational approaches to understand the adverse drug effect on potassium, sodium and calcium channels for predicting TdP cardiac arrhythmias
2017, Journal of Molecular Graphics and Modelling
Ion channels play a crucial role in the cardiovascular system. Our understanding of cardiac ion channel function has improved since their first discoveries. The flow of potassium, sodium and calcium ions across cardiomyocytes is vital for regular cardiac rhythm. Blockage of these channels, delays cardiac repolarization or tend to shorten repolarization and may induce arrhythmia. Detection of drug risk by channel blockade is considered essential for drug regulators. Advanced computational models can be used as an early screen for torsadogenic potential in drug candidates. New drug candidates that are determined to not cause blockage are more likely to pass successfully through preclinical trials and not be withdrawn later from the marketplace by manufacturer. Several different approved drugs, however, can cause a distinctive polymorphic ventricular arrhythmia known as torsade de pointes (TdP), which may lead to sudden death. The objective of the present study is to review the mechanisms and computational models used to assess the risk that a drug may TdP. Key points: There is strong evidence from multiple studies that blockage of the L-type calcium current reduces risk of TdP. Blockage of sodium channels slows cardiac action potential conduction, however, not all sodium channel blocking antiarrhythmic drugs produce a significant effect, while late sodium channel block reduces TdP. Interestingly, there are some drugs that block the hERG potassium channel and therefore cause QT prolongation, but they are not associated with TdP. Recent studies confirmed the necessity of studying multiple distinctionic ion channels which are responsible for cardiac related diseases or TdP, to obtain an improved clinical TdP risk prediction of compound interactions and also for designing drugs.
3D-SDAR modeling of hERG potassium channel affinity: A case study in model design and toxicophore identification
2017, Journal of Molecular Graphics and Modelling
A dataset of 237 human Ether-à-go-go Related Gene (hERG) potassium channel inhibitors (180 of which were used for model building and validation, whereas 57 constituted the “true” external prediction set) collected from 22 literature sources was modeled by 3D-SDAR. To produce reliable and reproducible classification models for hERG blocking, the initial set of 180 chemicals was split into two subsets: a balanced modeling set consisting of 118 compounds and an unbalanced validation set comprised of 62 compounds. A PLS bagging-like algorithm written in Matlab was used to process the data and assign each compound to one of the two (hERG+ or hERG-) activity classes. The best predictive model evaluated on the basis of a fully randomized hold-out test set (comprising 20% of the modeling set) used 4 latent variables and a grid of 6 ppm × 6 ppm × 1 Å in the C-C region, 6 ppm × 30 ppm × 1 Å in the C-N region, and 30 ppm × 30 ppm × 1 Å in the N-N region. An overall accuracy of 0.84 was obtained for both the hold-out test set and the validation set. Further, an external prediction set consisting of 57 drugs and drug derivatives was used to estimate the true predictive power of the reported 3D-SDAR model – a slight reduction of the overall accuracy down to 0.77 was observed. 3D-SDAR map of the most frequently occurring bins and their projection on the standard coordinate space of the chemical structures allowed identification of a three-center toxicophore composed of two aromatic rings and an amino group. A U test along the distance axis of the most frequently occurring 3D-SDAR bins was used to set the distance limits of the toxicophore. This toxicophore was found to be similar to an earlier reported phospholipidosis (PLD) toxicophore.
Monte Carlo method for predicting of cardiac toxicity: hERG blocker compounds
2016, Toxicology Letters
The estimation of the cardiotoxicity of compounds is an important task for the drug discovery as well as for the risk assessment in ecological aspect. The experimental estimation of the above endpoint is complex and expensive. Hence, the theoretical computational methods are very attractive alternative of the direct experiment. A model for cardiac toxicity of 400 hERG blocker compounds (pIC₅₀) is built up using the Monte Carlo method. Three different splits into the visible training set (in fact, the training set plus the calibration set) and invisible validation sets examined. The predictive potential is very good for all examined splits. The statistical characteristics for the external validation set are (i) the coefficient of determination r² = (0.90-0.93); and (ii) root-mean squared error s = (0.30–0.40).

View all citing articles on Scopus

View full text

Predictive models for hERG potassium channel blockers

Abstract

Graphical abstract

Section snippets

Additional material

Acknowledgments

Cell

J. Biomol. Screen.

Eur. J. Pharmacol.

J. Pharmacol. Sci.

Eur. J. Pharmacol.

Bioorg. Med. Chem. Lett.

Bioorg. Med. Chem.

Bioorg. Med. Chem. Lett.

J. Biol. Chem.

J. Biol. Chem.

Drug Safety

Curr. Opin. Drug Discov. Dev.

J. Med. Chem.

Curr. Opin. Drug Discov. Dev.