Chemical and biophysical characterization of novel potassium channel blocker 3-fluoro-5-methylpyridin-4-amine.

4-aminopyridine (4AP) is a potassium (K+) channel blocker used clinically to improve walking in people with multiple sclerosis (MS). 4AP binds to exposed K+ channels in demyelinated axons, reducing the leakage of intracellular K+ and enhancing impulse conduction. Multiple derivatives of 4AP capable of blocking K+ channels have been reported including three radiolabeled with positron emitting isotopes for imaging demyelinated lesions using positron emission tomography (PET). However, there remains a demand for novel molecules with suitable physicochemical properties and binding affinity that can potentially be radiolabeled and used as PET radiotracers. In this study, we introduce 3-fluoro-5-methylpyridin-4-amine (5Me3F4AP) as a novel trisubstituted K+ channel blocker with potential application in PET. 5Me3F4AP has comparable potency to 4AP and the PET tracer 3-fluoro-4-aminopyridine (3F4AP). Compared to 3F4AP, 5Me3F4AP exhibits comparable basicity (pKa = 7.46 ± 0.01 vs. 7.37 ± 0.07, P-value = 0.08), greater lipophilicity (logD = 0.664 ± 0.005 vs. 0.414 ± 0.002, P-value < 0.0001) and higher permeability to an artificial brain membrane (Pe = 88.1 ± 18.3 vs. 31.1 ± 2.9 nm/s, P-value = 0.03). 5Me3F4AP is also more stable towards oxidation in vitro by the cytochrome P450 enzyme CYP2E1 (IC50 = 36.2 ± 2.5 vs. 15.4 ± 5.1, P-value = 0.0003); the enzyme responsible for the metabolism of 4AP and 3F4AP. Taken together, 5Me3F4AP has promising properties as a candidate for PET imaging warranting additional investigation.

Synthetic ShK-like Peptide from the Jellyfish Nemopilema nomurai Has Human Voltage-Gated Potassium-Channel-Blocking Activity.

We identified a new human voltage-gated potassium channel blocker, NnK-1, in the jellyfish Nemopilema nomurai based on its genomic information. The gene sequence encoding NnK-1 contains 5408 base pairs, with five introns and six exons. The coding sequence of the NnK-1 precursor is 894 nucleotides long and encodes 297 amino acids containing five presumptive ShK-like peptides. An electrophysiological assay demonstrated that the fifth peptide, NnK-1, which was chemically synthesized, is an effective blocker of hKv1.3, hKv1.4, and hKv1.5. Multiple-sequence alignment with cnidarian Shk-like peptides, which have Kv1.3-blocking activity, revealed that three residues (3Asp, 25Lys, and 34Thr) of NnK-1, together with six cysteine residues, were conserved. Therefore, we hypothesized that these three residues are crucial for the binding of the toxin to voltage-gated potassium channels. This notion was confirmed by an electrophysiological assay with a synthetic peptide (NnK-1 mu) where these three peptides were substituted with 3Glu, 25Arg, and 34Met. In conclusion, we successfully identified and characterized a new voltage-gated potassium channel blocker in jellyfish that interacts with three different voltage-gated potassium channels. A peptide that interacts with multiple voltage-gated potassium channels has many therapeutic applications in various physiological and pathophysiological contexts.

The scorpion toxin BeKm-1 blocks hERG cardiac potassium channels using an indispensable arginine residue.

BeKm-1 is a peptide toxin from scorpion venom that blocks the pore of the potassium channel hERG (Kv11.1) in the human heart. Although individual protein structures have been resolved, the structure of the complex between hERG and BeKm-1 is unknown. Here, we used molecular dynamics and ensemble docking, guided by previous double-mutant cycle analysis data, to obtain an in silico model of the hERG-BeKm-1 complex. Adding to the previous mutagenesis study of BeKm-1, our model uncovers the key role of residue Arg20, which forms three interactions (a salt bridge and hydrogen bonds) with the channel vestibule simultaneously. Replacement of this residue even by lysine weakens the interactions significantly. In accordance, the recombinantly produced BeKm-1R20K mutant exhibited dramatically decreased activity on hERG. Our model may be useful for future drug design attempts.

KcsA-Kv1.x chimeras with complete ligand-binding sites provide improved predictivity for screening selective Kv1.x blockers.

Despite significant advances in the development of therapeutic interventions targeting autoimmune diseases and chronic inflammatory conditions, lack of effective treatment still poses a high unmet need. Modulating chronically activated T cells through the blockade of the Kv1.3 potassium channel is a promising therapeutic approach; however, developing selective Kv1.3 inhibitors is still an arduous task. Phage display-based high throughput peptide library screening is a rapid and robust approach to develop promising drug candidates; however, it requires solid-phase immobilization of target proteins with their binding site preserved. Historically, the KcsA bacterial channel chimera harboring only the turret region of the human Kv1.3 channel was used for screening campaigns. Nevertheless, literature data suggest that binding to this type of chimera does not correlate well with blocking potency on the native Kv1.3 channels. Therefore, we designed and successfully produced advanced KcsA-Kv1.3, KcsA-Kv1.1, and KcsA-Kv1.2 chimeric proteins in which both the turret and part of the filter regions of the human Kv1.x channels were transferred. These T+F (turret-filter) chimeras showed superior peptide ligand-binding predictivity compared to their T-only versions in novel phage ELISA assays. Phage ELISA binding and competition results supported with electrophysiological data confirmed that the filter region of KcsA-Kv1.x is essential for establishing adequate relative affinity order among selected peptide toxins (Vm24 toxin, Hongotoxin-1, Kaliotoxin-1, Maurotoxin, Stichodactyla toxin) and consequently obtaining more reliable selectivity data. These new findings provide a better screening tool for future drug development efforts and offer insight into the target-ligand interactions of these therapeutically relevant ion channels.

Isolation and structural identification of a potassium ion channel Kv4.1 inhibitor SsTx-P2 from centipede venom. / 蜈蚣毒液中钾离子通道Kv4.1抑制剂SsTx-P2的分离和结构鉴定.

OBJECTIVES: To isolate a potassium ion channel Kv4.1 inhibitor from centipede venom, and to determine its sequence and structure. METHODS: Ion-exchange chromatography and reversed-phase high-performance liquid chromatography were performed to separate and purify peptide components of centipede venom, and their inhibiting effect on Kv4.1 channel was determined by whole-cell patch clamp recording. The molecular weight of isolated peptide Kv4.1 channel inhibitor was identified with matrix assisted laser desorption ionization-time-of-flight mass spectrometry; its primary sequence was determined by Edman degradation sequencing and two-dimensional mass spectrometry; its structure was established based on iterative thread assembly refinement online analysis. RESULTS: A peptide SsTx-P2 was separated from centipede venom with the molecular weight of 6122.8, and its primary sequence consists of 53 amino acid residues NH2-ELTWDFVRTCCKLFPDKSECTKACATEFTGGDESRLKDVWPRKLRSGDSRLKD-OH. Peptide SsTx-P2 potently inhibited the current of Kv4.1 channel transiently transfected in HEK293 cell, with 1.0 µmol/L SsTx-P2 suppressing 95% current of Kv4.1 channel. Its structure showed that SsTx-P2 shared a conserved helical structure. CONCLUSIONS: The study has isolated a novel peptide SsTx-P2 from centipede venom, which can potently inhibit the potassium ion channel Kv4.1 and displays structural conservation.

Computational Studies Show How the H463R Mutation Turns hKv1.5 into an Inactivation State.

The KCNA5 gene provides the code for the α-subunit of the potassium channel Kv1.5. The genetic variant H463R in the Kv1.5 channel has been reported to cause a functional loss in atrial fibrillation (AF) patients. Understanding the mutations at a molecular level is key to developing improved therapeutics concerning cardiac hKv1.5 and hKv1.4 channels. Molecular dynamics and umbrella sampling free energy simulations are an effective tool to understand the mutation's effect on ion conduction, which we have employed and found that the hKv1.5[H463R] mutation imposes an energy barrier on the ion conduction pathway compared to the wild-type channel's ion free energy and pore structure. These results imply that the arginine mutation associated with the AF disease in particular modulates the inactivation process of hKv1.5. Kv1.4, encoded by the KCNA4 gene, is also present in the heart. Therefore, we considered simulation studies of the equivalent H507R mutation in the hKv1.4 channel and found that the mutation slightly reduces the ion conduction barrier in the ion conduction pathway, making it insignificant.

Benchmarking of Small Molecule Feature Representations for hERG, Nav1.5, and Cav1.2 Cardiotoxicity Prediction.

In the field of drug discovery, there is a substantial challenge in seeking out chemical structures that possess desirable pharmacological, toxicological, and pharmacokinetic properties. Complications arise when drugs interfere with the functioning of cardiac ion channels, leading to serious cardiovascular consequences. The discontinuation and removal of numerous approved drugs from the market or at late development stages in the pipeline due to such inhibitory effects further highlight the urgency of addressing this issue. Consequently, the early prediction of potential blockers targeting cardiac ion channels during the drug discovery process is of paramount importance. This study introduces a deep learning framework that computationally determines the cardiotoxicity associated with the voltage-gated potassium channel (hERG), the voltage-gated calcium channel (Cav1.2), and the voltage-gated sodium channel (Nav1.5) for drug candidates. The predictive capabilities of three feature representations─molecular fingerprints, descriptors, and graph-based numerical representations─are rigorously benchmarked. Additionally, a novel training and evaluation data set framework is presented, enabling predictive model training of drug off-target cardiotoxicity using a comprehensive and large curated data set covering these three cardiac ion channels. To facilitate these predictions, a robust and comprehensive small molecule cardiotoxicity prediction tool named CToxPred has been developed. It is made available as open source under the permissive MIT license at https://github.com/issararab/CToxPred.

Methionine-isoleucine dichotomy at a key position in scorpion toxins inhibiting voltage-gated potassium channels.

Previous studies have identified some key amino acid residues in scorpion toxins blocking potassium channels. In particular, the most numerous toxins belonging to the α-KTx family and affecting voltage-gated potassium channels (KV) present a conserved K-C-X-N motif in the C-terminal half of their sequence. Here, we show that the X position of this motif is almost always occupied by either methionine or isoleucine. We compare the activity of three pairs of peptides that differ just by this residue on a panel of KV1 channels and find that toxins bearing methionine affect preferentially KV1.1 and 1.6 isoforms. The refined K-C-M/I-N motif stands out as the principal structural element of α-KTx conferring high affinity and selectivity to KV channels.

Interaction of the Inhibitory Peptides ShK and HmK with the Voltage-Gated Potassium Channel KV1.3: Role of Conformational Dynamics.

Peptide toxins that adopt the ShK fold can inhibit the voltage-gated potassium channel KV1.3 with IC50 values in the pM range and are therefore potential leads for drugs targeting autoimmune and neuroinflammatory diseases. Nuclear magnetic resonance (NMR) relaxation measurements and pressure-dependent NMR have shown that, despite being cross-linked by disulfide bonds, ShK itself is flexible in solution. This flexibility affects the local structure around the pharmacophore for the KV1.3 channel blockade and, in particular, the relative orientation of the key Lys and Tyr side chains (Lys22 and Tyr23 in ShK) and has implications for the design of KV1.3 inhibitors. In this study, we have performed molecular dynamics (MD) simulations on ShK and a close homologue, HmK, to probe the conformational space occupied by the Lys and Tyr residues, and docked the different conformations with a recently determined cryo-EM structure of the KV1.3 channel. Although ShK and HmK have 60% sequence identity, their dynamic behaviors are quite different, with ShK sampling a broad range of conformations over the course of a 5 µs MD simulation, while HmK is relatively rigid. We also investigated the importance of conformational dynamics, in particular the distance between the side chains of the key dyad Lys22 and Tyr23, for binding to KV1.3. Although these peptides have quite different dynamics, the dyad in both adopts a similar configuration upon binding, revealing a conformational selection upon binding to KV1.3 in the case of ShK. Both peptides bind to KV1.3 with Lys22 occupying the pore of the channel. Intriguingly, the more flexible peptide, ShK, binds with significantly higher affinity than HmK.

Machine-learning technique, QSAR and molecular dynamics for hERG-drug interactions.

One of the most well-known anti-targets defining medication cardiotoxicity is the voltage-dependent hERG K + channel, which is well-known for its crucial involvement in cardiac action potential repolarization. Torsades de Pointes, QT prolongation, and sudden death are all caused by hERG (the human Ether-à-go-go-Related Gene) inhibition. There is great interest in creating predictive computational (in silico) tools to identify and weed out potential hERG blockers early in the drug discovery process because testing for hERG liability and the traditional experimental screening are complicated, expensive and time-consuming. This study used 2D descriptors of a large curated dataset of 6766 compounds and machine learning approaches to build robust descriptor-based QSAR and predictive classification models for KCNH2 liability. Decision Tree, Random Forest, Logistic Regression, Ada Boosting, kNN, SVM, Naïve Bayes, neural network and stochastic gradient classification classifier algorithms were used to build classification models. If a compound's IC50 value was between 10 µM and less, it was classified as a blocker (hERG-positive), and if it was more, it was classified as a non-blocker (hERG-negative). Matthew's correlation coefficient formula and F1score were applied to compare and track the developed models' performance. Molecular docking and dynamics studies were performed to understand the cardiotoxicity relating to the hERG-gene. The hERG residues interacting after 100 ns are LEU:697, THR:708, PHE:656, HIS:674, HIS:703, TRP:705 and ASN:709 and the hERG-ligand-16 complex trajectory showed stable behaviour with lesser fluctuations in the entire simulation of 200 ns.Communicated by Ramaswamy H. Sarma.

AgTx2-GFP, Fluorescent Blocker Targeting Pharmacologically Important Kv1.x (x = 1, 3, 6) Channels.

The growing interest in potassium channels as pharmacological targets has stimulated the development of their fluorescent ligands (including genetically encoded peptide toxins fused with fluorescent proteins) for analytical and imaging applications. We report on the properties of agitoxin 2 C-terminally fused with enhanced GFP (AgTx2-GFP) as one of the most active genetically encoded fluorescent ligands of potassium voltage-gated Kv1.x (x = 1, 3, 6) channels. AgTx2-GFP possesses subnanomolar affinities for hybrid KcsA-Kv1.x (x = 3, 6) channels and a low nanomolar affinity to KcsA-Kv1.1 with moderate dependence on pH in the 7.0-8.0 range. Electrophysiological studies on oocytes showed a pore-blocking activity of AgTx2-GFP at low nanomolar concentrations for Kv1.x (x = 1, 3, 6) channels and at micromolar concentrations for Kv1.2. AgTx2-GFP bound to Kv1.3 at the membranes of mammalian cells with a dissociation constant of 3.4 ± 0.8 nM, providing fluorescent imaging of the channel membranous distribution, and this binding depended weakly on the channel state (open or closed). AgTx2-GFP can be used in combination with hybrid KcsA-Kv1.x (x = 1, 3, 6) channels on the membranes of E. coli spheroplasts or with Kv1.3 channels on the membranes of mammalian cells for the search and study of nonlabeled peptide pore blockers, including measurement of their affinity.

Structural analysis of hERG channel blockers and the implications for drug design.

The repolarizing current (Ikr) produced by the hERG potassium channel forms a major component of the cardiac action potential and blocking this current by small molecule drugs can lead to life-threatening cardiotoxicity. Understanding the mechanisms of drug-mediated hERG inhibition is essential to develop a second generation of safe drugs, with minimal cardiotoxic effects. Although various computational tools and drug design guidelines have been developed to avoid binding of drugs to the hERG pore domain, there are many other aspects that are still open for investigation. This includes the use computational modelling to study the implications of hERG mutations on hERG structure and trafficking, the interactions of hERG with hERG chaperone proteins and with membrane-soluble molecules, the mechanisms of drugs that inhibit hERG trafficking and drugs that rescue hERG mutations. The plethora of available experimental data regarding all these aspects can guide the construction of much needed robust computational structural models to study these mechanisms for the rational design of safe drugs.

Characterization and Chemical Synthesis of Cm39 (α-KTx 4.8): A Scorpion Toxin That Inhibits Voltage-Gated K+ Channel KV1.2 and Small- and Intermediate-Conductance Ca2+-Activated K+ Channels KCa2.2 and KCa3.1.

A novel peptide, Cm39, was identified in the venom of the scorpion Centruroides margaritatus. Its primary structure was determined. It consists of 37 amino acid residues with a MW of 3980.2 Da. The full chemical synthesis and proper folding of Cm39 was obtained. Based on amino acid sequence alignment with different K+ channel inhibitor scorpion toxin (KTx) families and phylogenetic analysis, Cm39 belongs to the α-KTx 4 family and was registered with the systematic number of α-KTx 4.8. Synthetic Cm39 inhibits the voltage-gated K+ channel hKV1.2 with high affinity (Kd = 65 nM). The conductance-voltage relationship of KV1.2 was not altered in the presence of Cm39, and the analysis of the toxin binding kinetics was consistent with a bimolecular interaction between the peptide and the channel; therefore, the pore blocking mechanism is proposed for the toxin-channel interaction. Cm39 also inhibits the Ca2+-activated KCa2.2 and KCa3.1 channels, with Kd = 502 nM, and Kd = 58 nM, respectively. However, the peptide does not inhibit hKV1.1, hKV1.3, hKV1.4, hKV1.5, hKV1.6, hKV11.1, mKCa1.1 K+ channels or the hNaV1.5 and hNaV1.4 Na+ channels at 1 µM concentrations. Understanding the unusual selectivity profile of Cm39 motivates further experiments to reveal novel interactions with the vestibule of toxin-sensitive channels.

Virtual screening of DrugBank database for hERG blockers using topological Laplacian-assisted AI models.

The human ether-a-go-go (hERG) potassium channel (Kv11.1) plays a critical role in mediating cardiac action potential. The blockade of this ion channel can potentially lead fatal disorder and/or long QT syndrome. Many drugs have been withdrawn because of their serious hERG-cardiotoxicity. It is crucial to assess the hERG blockade activity in the early stage of drug discovery. We are particularly interested in the hERG-cardiotoxicity of compounds collected in the DrugBank database considering that many DrugBank compounds have been approved for therapeutic treatments or have high potential to become drugs. Machine learning-based in silico tools offer a rapid and economical platform to virtually screen DrugBank compounds. We design accurate and robust classifiers for blockers/non-blockers and then build regressors to quantitatively analyze the binding potency of the DrugBank compounds on the hERG channel. Molecular sequences are embedded with two natural language processing (NLP) methods, namely, autoencoder and transformer. Complementary three-dimensional (3D) molecular structures are embedded with two advanced mathematical approaches, i.e., topological Laplacians and algebraic graphs. With our state-of-the-art tools, we reveal that 227 out of the 8641 DrugBank compounds are potential hERG blockers, suggesting serious drug safety problems. Our predictions provide guidance for the further experimental interrogation of DrugBank compounds' hERG-cardiotoxicity.

Investigating cardiotoxicity related with hERG channel blockers using molecular fingerprints and graph attention mechanism.

Human ether-a-go-go-related gene (hERG) channel blockade by small molecules is a big concern during drug development in the pharmaceutical industry. Failure or inhibition of hERG channel activity caused by drug molecules can lead to prolonging QT interval, which will result in serious cardiotoxicity. Thus, evaluating the hERG blocking activity of all these small molecular compounds is technically challenging, and the relevant procedures are expensive and time-consuming. In this study, we develop a novel deep learning predictive model named DMFGAM for predicting hERG blockers. In order to characterize the molecule more comprehensively, we first consider the fusion of multiple molecular fingerprint features to characterize its final molecular fingerprint features. Then, we use the multi-head attention mechanism to extract the molecular graph features. Both molecular fingerprint features and molecular graph features are fused as the final features of the compounds to make the feature expression of compounds more comprehensive. Finally, the molecules are classified into hERG blockers or hERG non-blockers through the fully connected neural network. We conduct 5-fold cross-validation experiment to evaluate the performance of DMFGAM, and verify the robustness of DMFGAM on external validation datasets. We believe DMFGAM can serve as a powerful tool to predict hERG channel blockers in the early stages of drug discovery and development.

Insight on the interaction between the scorpion toxin blocker Discrepin on potassium voltage-gated channel Kv4.3 by molecular dynamics simulations.

Discrepin is a 38-residue α-toxin extracted from the venom of the Venezuelan scorpion Tityus discrepans, which inhibits ionic transit in the voltage-dependent potassium channels (Kv) of A-type current. The effect of specific residues on the IC50 between Discrepine and Kv4.3, the main component of A-type currents, is known; however, the molecular details of the toxin-channel interaction are not known. In this work, we present interaction models between Discrepin (wt) and two peptide variants (V6K/D20K and K13A) on the pore-forming domain of the Kv4.3 channel obtained from homology, docking, and molecular dynamics modeling techniques. The free energy calculations in these models correspond to the order of the experimentally determined IC50 values. Our studies shed light on the role of the K13 residue as responsible for occluding the Kv4.3 selectivity filter and the importance of the V6K mutation in the approach and stabilization of toxin-channel complex interactions.Communicated by Ramaswamy H. Sarma.

Combining mKate2-Kv1.3 Channel and Atto488-Hongotoxin for the Studies of Peptide Pore Blockers on Living Eukaryotic Cells.

The voltage-gated potassium Kv1.3 channel is an essential component of vital cellular processes which is also involved in the pathogenesis of some autoimmune, neuroinflammatory and oncological diseases. Pore blockers of the Kv1.3 channel are considered as potential drugs and are used to study Kv1 channels' structure and functions. Screening and study of the blockers require the assessment of their ability to bind the channel. Expanding the variety of methods used for this, we report on the development of the fluorescent competitive binding assay for measuring affinities of pore blockers to Kv1.3 at the membrane of mammalian cells. The assay constituents are hongotoxin 1 conjugated with Atto488, fluorescent mKate2-tagged Kv1.3 channel, which was designed to improve membrane expression of the channel in mammalian cells, confocal microscopy, and a special protocol of image processing. The assay is implemented in the "mix and measure", format and allows the screening of Kv1.3 blockers, such as peptide toxins, that bind to the extracellular vestibule of the K+-conducting pore, and analyzing their affinity.

A Fluorescent Peptide Toxin for Selective Visualization of the Voltage-Gated Potassium Channel KV1.3.

Upregulation of the voltage-gated potassium channel KV1.3 is implicated in a range of autoimmune and neuroinflammatory diseases, including rheumatoid arthritis, psoriasis, multiple sclerosis, and type I diabetes. Understanding the expression, localization, and trafficking of KV1.3 in normal and disease states is key to developing targeted immunomodulatory therapies. HsTX1[R14A], an analogue of a 34-residue peptide toxin from the scorpion Heterometrus spinifer, binds KV1.3 with high affinity (IC50 of 45 pM) and selectivity (2000-fold for KV1.3 over KV1.1). We have synthesized a fluorescent analogue of HsTX1[R14A] by N-terminal conjugation of a Cy5 tag. Electrophysiology assays show that Cy5-HsTX1[R14A] retains activity against KV1.3 (IC50 ∼ 0.9 nM) and selectivity over a range of other potassium channels (KV1.2, KV1.4, KV1.5, KV1.6, KCa1.1 and KCa3.1), as well as selectivity against heteromeric channels assembled from KV1.3/KV1.5 tandem dimers. Live imaging of CHO cells expressing green fluorescent protein-tagged KV1.3 shows co-localization of Cy5-HsTX1[R14A] and KV1.3 fluorescence signals at the cell membrane. Moreover, flow cytometry demonstrated that Cy5-HsTX1[R14A] can detect KV1.3-expressing CHO cells. Stimulation of mouse microglia by lipopolysaccharide, which enhances membrane expression of KV1.3, was associated with increased staining by Cy5-HsTX1[R14A], demonstrating that it can be used to identify KV1.3 in disease-relevant models of inflammation. Furthermore, the biodistribution of Cy5-HsTX1[R14A] could be monitored using ex vivo fluorescence imaging of organs in mice dosed subcutaneously with the peptide. These results illustrate the utility of Cy5-HsTX1[R14A] as a tool for visualizing KV1.3, with broad applicability in fundamental investigations of KV1.3 biology, and the validation of novel disease indications where KV1.3 inhibition may be of therapeutic value.

Physicochemical QSAR analysis of hERG inhibition revisited: towards a quantitative potency prediction.

In an earlier study (Didziapetris R & Lanevskij K (2016). J Comput Aided Mol Des. 30:1175-1188) we collected a database of publicly available hERG inhibition data for almost 6700 drug-like molecules and built a probabilistic Gradient Boosting classifier with a minimal set of physicochemical descriptors (log P, pKa, molecular size and topology parameters). This approach favored interpretability over statistical performance but still achieved an overall classification accuracy of 75%. In the current follow-up work we expanded the database (provided in Supplementary Information) to almost 9400 molecules and performed temporal validation of the model on a set of novel chemicals from recently published lead optimization projects. Validation results showed almost no performance degradation compared to the original study. Additionally, we rebuilt the model using AFT (Accelerated Failure Time) learning objective in XGBoost, which accepts both quantitative and censored data often reported in protein inhibition studies. The new model achieved a similar level of accuracy of discerning hERG blockers from non-blockers at 10 µM threshold, which can be conceived as close to the performance ceiling for methods aiming to describe only non-specific ligand interactions with hERG. Yet, this model outputs quantitative potency values (IC50) and is not tied to a particular classification cut-off. pIC50 from patch-clamp measurements can be predicted with R2 ≈ 0.4 and MAE < 0.5, which enables ligand ranking according to their expected potency levels. The employed approach can be valuable for quantitative modeling of various ADME and drug safety endpoints with a high prevalence of censored data.

Artificial pore blocker acts specifically on voltage-gated potassium channel isoform KV1.6.

Among voltage-gated potassium channel (KV) isoforms, KV1.6 is one of the most widespread in the nervous system. However, there are little data concerning its physiological significance, in part due to the scarcity of specific ligands. The known high-affinity ligands of KV1.6 lack selectivity, and conversely, its selective ligands show low affinity. Here, we present a designer peptide with both high affinity and selectivity to KV1.6. Previously, we have demonstrated that KV isoform-selective peptides can be constructed based on the simplistic α-hairpinin scaffold, and we obtained a number of artificial Tk-hefu peptides showing selective blockage of KV1.3 in the submicromolar range. We have now proposed amino acid substitutions to enhance their activity. As a result, we have been able to produce Tk-hefu-11 that shows an EC50 of ≈70 nM against KV1.3. Quite surprisingly, Tk-hefu-11 turns out to block KV1.6 with even higher potency, presenting an EC50 of ≈10 nM. Furthermore, we have solved the peptide structure and used molecular dynamics to investigate the determinants of selective interactions between artificial α-hairpinins and KV channels to explain the dramatic increase in KV1.6 affinity. Since KV1.3 is not highly expressed in the nervous system, we hope that Tk-hefu-11 will be useful in studies of KV1.6 and its functions.