Volume 8 Supplement 8
Homology modeling of a voltage-gated potassium channel (human Kv7.1)
© Goncearenco et al; licensee BioMed Central Ltd. 2007
Published: 20 November 2007
Long QT Syndrome (LQTS) is a heart disease in which the QT interval of the electrocardiogram, required for repolarization of the ventricles in the heart, is abnormally long. It is associated with syncope, arrhythmias, torsade de pointes and sudden cardiac death due to ventricular fibrillation. LQT1, the most common type of LQTS, is caused by mutations in the KCNQ1 gene. The product of this gene is the Kv7.1 (K+ channel voltage-dependent variant 7.1) protein. The Kv7.1 channel mediates the slowly activating delayed rectifier potassium current (IKs) which in turn is one of the two currents responsible for ventricular depolarization in the heart. The Kv7.1 α subunit co-assembles with the minK potassium channel β subunit (coded by the KCNE1 gene) with a α4/β4 stoichiometry to form the fully functional ion channel.
Materials and methods
This study focuses on the elucidation of the 3D structure of the human Kv7.1 channel using a homology modeling approach based on the mammalian Kv1.2  and bacterial KvAP (potassium channel-voltage gated – Aeropyrum Pernix)  crystallographic structure templates, which represent different gating states of the channel, i.e. intermediate vs. fully open. The Schrödinger molecular modeling software suite was used to build and refine the models. High emphasis was placed on sequence alignment, topology, restrained energy minimization and maintenance of symmetry.
Results and discussion
Validation of the two resulting homology models was carried out by comparison with experimental mutagenesis data. In particular, mutations in residues in the Kv7.1 channel that altered the potency of block by L-735821 (L-7), a high affinity benzodiazepine blocker of wild-type Kv7.1 channels , were observed to be in close proximity to L-7 docked into the inner channel vestibule in the Kv7.1 homology model, in good agreement with the experimentally determined results. Also, the observed contacts between the pore helix and transmembrane segments S5 and S6 were shown to be in agreement with experimental mutagenesis studies of co-localized residues . The model based on the KvAP template is in closer agreement with the experimental observations. Accessibility of the gating site, as observed in the Kv1.2-based model, suggests putative KCNQ1-KCNE1 inter-subunit contacts and allosteric modulation of the IKs channel, which is also in agreement with the available experimental data .
To estimate the ability of the models to predict blocker potency, virtual screening was performed and correlations between experimental pIC50s and docking scores were calculated. A literature search was carried out to collect a sufficient number of blockers with known inhibitory effects on the IKs channel, particularly IC50 values from patch-clamp experiments. None of the two models performed particularly well in these virtual screening runs. Although it is at present unclear which factors – apart from the known weaknesses of present-day scoring functions – may have caused these disappointing results, we hypothesize that the conformational state of the channel model will have a significant effect on the docking results. The reason is that channel gating produces a substantial change in the volume of the inner vestibule, which is believed to be the principle binding site for small molecule channel blockers. In order to define conformations of templates and models with respect to gating, the swink (swivel and kink) algorithm , which analyzes hinges in the gating helices, was implemented as a PyMol plug-in. This study suggests that in the absence of mammalian channel structures in the fully open state, homology building of potassium channel models may profit from introducing conformational changes into structures of most closely related templates first, in order to improve blockade predictions. On the other hand, physicochemical properties of ligands, such as molecular weight and log D also have to be considered as factors affecting the blockade and the voltage clamping results.
This effort facilitates the development of predictive pharmacophore models for the in silico screening of new compounds as a part of cardiotoxicity tests.
- Long SB, Campbell EB, Mackinnon R: Crystal structure of a mammalian voltage-dependent shaker family K+ channel. Science 2005, 309: 897–903. 10.1126/science.1116269View ArticlePubMedGoogle Scholar
- Jiang Y, Lee A, Chen J, Ruta V, Cadene M, Chait BT, MacKinnon R: X-ray structure of a voltage-dependent K+ channel. Nature 2003, 423: 33–41. 10.1038/nature01580View ArticlePubMedGoogle Scholar
- Seebohm G, Chen J, Strutz N, Culberson C, Lerche C, Sanguinetti MC: Molecular determinants of KCNQ1 channel block by a benzodiazepine. Mol Pharmacol 2003, 64: 70–77. 10.1124/mol.64.1.70View ArticlePubMedGoogle Scholar
- Seebohm G, Westenskow P, Lang F, Sanguinetti MC: Mutation of colocalized residues of the pore helix and transmembrane segments S5 and S6 disrupt deactivation and modify inactivation of KCNQ1 K+ channels. J Physiol 2005, 563: 359–368. 10.1113/jphysiol.2004.080887PubMed CentralView ArticlePubMedGoogle Scholar
- Kurokawa J, Motoike HK, Kass RS: Tea(+)-sensitive KCNQ1 constructs reveal pore-independent access to KCNE1 in assembled I(Ks) channels. J Gen Physiol 2001, 117: 43–52. 10.1085/jgp.117.1.43PubMed CentralView ArticlePubMedGoogle Scholar
- Sands Z, Grottesi A, Sansom MSP: Voltage-gated ion channels. Curr Biol 2005, 15: R44-R47. 10.1016/j.cub.2004.12.050View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd.