Voltar
Artigo Acesso aberto Revisado por pares
BIBTEX RIS

Molecular mechanism of EAG1 channel inhibition by imipramine binding to the PAS domain

2023; Elsevier BV; Volume: 299; Issue: 12 Linguagem: Inglês

10.1016/j.jbc.2023.105391

ISSN

1083-351X

Autores

Zejun Wang, Mahdi Ghorbani, Xi Chen, Purushottam B. Tiwari, Jeffery B. Klauda, Tinatin I. Brelidze,

Tópico(s)

Neuroscience and Neuropharmacology Research

Resumo

Ether-a-go-go (EAG) channels are key regulators of neuronal excitability and tumorigenesis. EAG channels contain an N-terminal Per-Arnt-Sim (PAS) domain that can regulate currents from EAG channels by binding small molecules. The molecular mechanism of this regulation is not clear. Using surface plasmon resonance and electrophysiology we show that a small molecule ligand imipramine can bind to the PAS domain of EAG1 channels and inhibit EAG1 currents via this binding. We further used a combination of molecular dynamics (MD) simulations, electrophysiology, and mutagenesis to investigate the molecular mechanism of EAG1 current inhibition by imipramine binding to the PAS domain. We found that Tyr71, located at the entrance to the PAS domain cavity, serves as a "gatekeeper" limiting access of imipramine to the cavity. MD simulations indicate that the hydrophobic electrostatic profile of the cavity facilitates imipramine binding and in silico mutations of hydrophobic cavity-lining residues to negatively charged glutamates decreased imipramine binding. Probing the PAS domain cavity-lining residues with site-directed mutagenesis, guided by MD simulations, identified D39 and R84 as residues essential for the EAG1 channel inhibition by imipramine binding to the PAS domain. Taken together, our study identified specific residues in the PAS domain that could increase or decrease EAG1 current inhibition by imipramine binding to the PAS domain. These findings should further the understanding of molecular mechanisms of EAG1 channel regulation by ligands and facilitate the development of therapeutic agents targeting these channels. Ether-a-go-go (EAG) channels are key regulators of neuronal excitability and tumorigenesis. EAG channels contain an N-terminal Per-Arnt-Sim (PAS) domain that can regulate currents from EAG channels by binding small molecules. The molecular mechanism of this regulation is not clear. Using surface plasmon resonance and electrophysiology we show that a small molecule ligand imipramine can bind to the PAS domain of EAG1 channels and inhibit EAG1 currents via this binding. We further used a combination of molecular dynamics (MD) simulations, electrophysiology, and mutagenesis to investigate the molecular mechanism of EAG1 current inhibition by imipramine binding to the PAS domain. We found that Tyr71, located at the entrance to the PAS domain cavity, serves as a "gatekeeper" limiting access of imipramine to the cavity. MD simulations indicate that the hydrophobic electrostatic profile of the cavity facilitates imipramine binding and in silico mutations of hydrophobic cavity-lining residues to negatively charged glutamates decreased imipramine binding. Probing the PAS domain cavity-lining residues with site-directed mutagenesis, guided by MD simulations, identified D39 and R84 as residues essential for the EAG1 channel inhibition by imipramine binding to the PAS domain. Taken together, our study identified specific residues in the PAS domain that could increase or decrease EAG1 current inhibition by imipramine binding to the PAS domain. These findings should further the understanding of molecular mechanisms of EAG1 channel regulation by ligands and facilitate the development of therapeutic agents targeting these channels. Ether-a-go-go (EAG) channels, also known as Kv10.1 and KCNH1 channels, contribute to the regulation of neuronal excitability (1Cázares-Ordoñez V. Pardo L.A. Kv10.1 potassium channel: from the brain to the tumors.Biochem. Cell Biol. 2017; 95: 531-536Crossref PubMed Scopus (32) Google Scholar, 2Bauer C.K. Schwarz J.R. Ether-à-go-go K+ channels: effective modulators of neuronal excitability.J. Physiol. 2018; 596: 769-783Crossref PubMed Scopus (44) Google Scholar, 3Kessi M. Chen B. Peng J. Tang Y. Olatoutou E. He F. et al.Intellectual disability and potassium channelopathies: a systematic review.Front. Genet. 2020; 11: 614Crossref PubMed Scopus (22) Google Scholar) and gain-of-function genetic mutations in EAG channels are associated with severe neurological disorders (4Fukai R. Saitsu H. Tsurusaki Y. Sakai Y. Haginoya K. Takahashi K. et al.De novo KCNH1 mutations in four patients with syndromic developmental delay, hypotonia and seizures.J. Hum. Genet. 2016; 61: 381-387Crossref PubMed Scopus (38) Google Scholar, 5Mastrangelo M. Scheffer I.E. Bramswig N.C. Nair L.D. Myers C.T. Dentici M.L. et al.Epilepsy in KCNH1-related syndromes.Epileptic Disord. 2016; 18: 123-136Crossref PubMed Scopus (31) Google Scholar, 6Veeramah K.R. Johnstone L. Karafet T.M. Wolf D. Sprissler R. Salogiannis J. et al.Exome sequencing reveals new causal mutations in children with epileptic encephalopathies.Epilepsia. 2013; 54: 1270-1281Crossref PubMed Scopus (211) Google Scholar, 7Kortüm F. Caputo V. Bauer C.K. Stella L. Ciolfi A. Alawi M. et al.Mutations in KCNH1 and ATP6V1B2 cause Zimmermann-Laband syndrome.Nat. Genet. 2015; 47: 661-667Crossref PubMed Scopus (156) Google Scholar, 8Simons C. Rash L.D. Crawford J. Ma L. Cristofori-Armstrong B. Miller D. et al.Mutations in the voltage-gated potassium channel gene KCNH1 cause Temple-Baraitser syndrome and epilepsy.Nat. Genet. 2015; 47: 73-77Crossref PubMed Scopus (108) Google Scholar). In addition to their function in the brain, EAG channels are upregulated in cancer (9Farias L.M. Ocańa D.B. Díaz L. Larrea F. Avila-Chávez E. Cadena A. et al.Ether á go-go potassium channels as human cervical cancer markers.Cancer Res. 2004; 64: 6996-7001Crossref PubMed Scopus (127) Google Scholar, 10Hemmerlein B. Weseloh R.M. Mello de Queiroz F. Knotgen H. Sanchez A. Rubio M.E. et al.Overexpression of Eag1 potassium channels in clinical tumours.Mol. Cancer. 2006; 5: 41Crossref PubMed Scopus (217) Google Scholar, 11Pardo L.A. Del C.D. Sanchez A. Alves F. Bruggemann A. Beckh S. et al.Oncogenic potential of EAG K+ channels.EMBO J. 1999; 18: 5540-5547Crossref PubMed Scopus (369) Google Scholar, 12Pardo L.A. Stühmer W. The roles of K+ channels in cancer.Nat. Rev. Cancer. 2014; 14: 39-48Crossref PubMed Scopus (336) Google Scholar, 13Rao V.R. Perez-Neut M. Kaja S. Gentile S. Voltage-gated ion channels in cancer cell proliferation.Cancers (Basel). 2015; 7: 849-875Crossref PubMed Scopus (135) Google Scholar) and inhibition of their activity has been shown to reduce cancer cell proliferation (14Gómez-Varela D. Zwick-Wallasch E. Knötgen H. Sánchez A. Hettmann T. Ossipov D. et al.Monoclonal antibody blockade of the human Eag1 potassium channel function exerts antitumor activity.Cancer Res. 2007; 67: 7343-7349Crossref PubMed Scopus (176) Google Scholar, 15Hartung F. Stühmer W. Pardo L.A. Tumor cell-selective apoptosis induction through targeting of KV10.1 via bifunctional TRAIL antibody.Mol. Cancer. 2011; 10: 109Crossref PubMed Scopus (50) Google Scholar, 16Weber C. de Queiroz F.M. Downie B.R. Suckow A. Stühmer W. Pardo L.A. Silencing the activity and proliferative properties of the human EagI potassium channel by RNA interference.J. Biol. Chem. 2006; 281: 13030-13037Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar). EAG channels belong to the KCNH family of voltage-gated potassium channels that also include EAG-related gene (ERG) and EAG-like K+ (ELK) subfamilies (17Ganetzky B. Robertson G.A. Wilson G.F. Trudeau M.C. Titus S.A. The Eag family of K+ channels in drosophila and mammals.Ann. N. Y. Acad. Sci. 1999; 868: 356-369Crossref PubMed Scopus (68) Google Scholar, 18Warmke J.W. Ganetzky B. A family of potassium channel genes related to Eag in Drosophila and mammals.Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 3438-3442Crossref PubMed Google Scholar). Similar to other voltage-gated potassium channels, EAG channels are assembled from four subunits, each containing six membrane-spanning segments (S1–S6) (17Ganetzky B. Robertson G.A. Wilson G.F. Trudeau M.C. Titus S.A. The Eag family of K+ channels in drosophila and mammals.Ann. N. Y. Acad. Sci. 1999; 868: 356-369Crossref PubMed Scopus (68) Google Scholar, 19Wang W. MacKinnon R. Cryo-EM structure of the open human Ether-à-go-go -related K + channel hERG.Cell. 2017; 169: 422-430Abstract Full Text Full Text PDF PubMed Scopus (345) Google Scholar, 20Whicher J.R. MacKinnon R. Structure of the voltage-gated K + channel Eag1 reveals an alternative voltage sensing mechanism.Science. 2016; 353: 664-669Crossref PubMed Scopus (226) Google Scholar). The S1–S4 segments form a voltage-sensor domain (VSD) and segments S5–S6 form a pore-domain (PD) (Fig. 1A). The centrally located pore of the channel is formed from the PDs of all four subunits, with the p-loops between the S5 and S6 segments lining the selectivity filter of the channel. Each subunit of EAG channels also contains an N-terminal Per-Arnt-Sim (PAS) domain and a C-terminal cyclic nucleotide-binding homology (CNBH) domain (17Ganetzky B. Robertson G.A. Wilson G.F. Trudeau M.C. Titus S.A. The Eag family of K+ channels in drosophila and mammals.Ann. N. Y. Acad. Sci. 1999; 868: 356-369Crossref PubMed Scopus (68) Google Scholar, 20Whicher J.R. MacKinnon R. Structure of the voltage-gated K + channel Eag1 reveals an alternative voltage sensing mechanism.Science. 2016; 353: 664-669Crossref PubMed Scopus (226) Google Scholar, 21Morais Cabral J.H. Lee A. Cohen S.L. Chait B.T. Li M. MacKinnon R. Crystal structure and functional analysis of the HERG potassium channel N terminus: a eukaryotic PAS domain.Cell. 1998; 95: 649-655Abstract Full Text Full Text PDF PubMed Google Scholar). The PAS and CNBH domains from adjacent subunits form inter-subunit interactions in EAG channels (20Whicher J.R. MacKinnon R. Structure of the voltage-gated K + channel Eag1 reveals an alternative voltage sensing mechanism.Science. 2016; 353: 664-669Crossref PubMed Scopus (226) Google Scholar, 22Gustina A.S. Trudeau M.C. A recombinant N-terminal domain fully restores deactivation gating in N-truncated and long QT syndrome mutant hERG potassium channels.Proc. Natl. Acad. Sci. U. S. A. 2009; 106: 13082-13087Crossref PubMed Scopus (77) Google Scholar, 23Haitin Y. Carlson A.E. Zagotta W.N. The structural mechanism of KCNH-channel regulation by the Eag domain.Nature. 2013; 501: 444-448Crossref PubMed Scopus (88) Google Scholar). In addition to forming the intersubunit interactions, it has been shown that PAS domains of EAG channels bind small molecule ligands chlorpromazine and undecylenic acid (24Wang Z.J. Tiwari P.B. Üren A. Brelidze T.I. Identification of undecylenic acid as EAG channel inhibitor using surface plasmon resonance-based screen of KCNH channels.BMC Pharmacol. Toxicol. 2019; 20: 42Crossref PubMed Scopus (4) Google Scholar, 25Wang Z.J. Soohoo S.M. Tiwari P.B. Piszczek G. Brelidze T.I. Chlorpromazine binding to the PAS domains uncovers the effect of ligand modulation on EAG channel activity.J. Biol. Chem. 2020; 295: 4114-4123Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar), and PAS domains of related ERG3 channels bind heme (26Burton M.J. Cresser-Brown J. Thomas M. Portolano N. Basran J. Freeman S.L. et al.Discovery of a heme-binding domain in a neuronal voltage-gated potassium channel.J. Biol. Chem. 2020; 295: 13277-13286Abstract Full Text Full Text PDF PubMed Scopus (0) Google Scholar). While it has been proposed that these small molecule ligands cause inhibition of currents through these channels via the binding to the PAS domains, so far, the only evidence in support of this mechanism is that the deletion of the PAS domain decreases the inhibition of EAG currents by chlorpromazine (25Wang Z.J. Soohoo S.M. Tiwari P.B. Piszczek G. Brelidze T.I. Chlorpromazine binding to the PAS domains uncovers the effect of ligand modulation on EAG channel activity.J. Biol. Chem. 2020; 295: 4114-4123Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar). However, deletion of an entire domain could have unforeseen effects on the channel gating that could mask the mechanism of the small molecules effect on EAG and related channels. Here we identified, for the first time, specific residues on the PAS domain of EAG1 channels that facilitate or limit EAG current inhibition by small molecule ligands. We used surface plasmon resonance (SPR) to show that a small molecule ligand imipramine directly binds to the PAS domain of EAG1 channels. We then used electrophysiology and site-directed mutagenesis, guided by molecular dynamics (MD) simulations, to further examine the molecular mechanism of EAG1 current inhibition by imipramine. We found that deletion of the PAS domain substantially decreased EAG1 current inhibition by imipramine, suggesting that the effect of imipramine is, at least in part, mediated via binding to the PAS domain. Docking of imipramine into the PAS domain cavity and accompanied MD simulations identified Tyr71, located at the entrance to the cavity, as a "gatekeeper" residue, limiting access of imipramine to the binding site inside the cavity. Consistent with this, substitution of Tyr71 with glycine and valine increased EAG1 current inhibition by imipramine, while substitution with phenylalanine had no effect on the current inhibition by imipramine. MD simulations indicated that the hydrophobic profile of the cavity facilitates the binding of imipramine and in silico mutations of four PAS domain lining hydrophobic residues to negatively charged glutamates decreased imipramine binding. Computational modeling-driven mutagenesis studies of residues lining the PAS domain cavity identified residues D39 and R84 as essential for the EAG1 current inhibition by imipramine binding to the PAS domain. Consistent with this, substitutions of D39 and R84 residues with glycine substantially decreased EAG current inhibition by imipramine. The residual inhibition of EAG1 channels lacking the PAS domain, combined with the previously reported inhibition of EAG1 channels by imipramine via an open-pore block (27García-Ferreiro R.E. Kerschensteiner D. Major F. Monje F. Stühmer W. Pardo L.A. Mechanism of block of hEag1 K+ channels by imipramine and astemizole.J. Gen. Physiol. 2004; 124: 301-317Crossref PubMed Scopus (115) Google Scholar), indicates that imipramine inhibits EAG1 currents by a dual mechanism: by binding to the PAS domain and by blocking the conduction pore. Taken together, our results shed light on the molecular mechanism of EAG1 current inhibition by the PAS domain small molecule binders. To test if imipramine can directly bind to the PAS domain of EAG1 channels we immobilized purified isolated PAS domains of EAG1 channels on the CM5 sensor ship using amine coupling and determined the SPR response over the range of concentrations of imipramine injected over the immobilized PAS domains. The SPR response increased with the increase in the imipramine concentration (Fig. 1B). Because of non-specific binding of imipramine to the CM5 sensor chip at high concentrations, we were unable to obtain the complete concentration dependence of the binding necessary to determine the binding affinity. However, the concentration-dependent increase in the SPR response for the tested imipramine concentrations indicates that similar to chlorpromazine, imipramine directly binds to the PAS domain of EAG1 channels. To determine the functional effect of imipramine on EAG1 channels we recorded currents from EAG1 channels using the two-electrode voltage clamp (TEVC) technique in the absence and presence of imipramine. Imipramine inhibited EAG1 currents in a concentration-dependent manner with the IC50 of 58.1 ± 9.7 μM at +50 mV (Fig. 2, A and B, and Table 1). To further determine the contribution of imipramine binding to the PAS domain to the current inhibition, currents were recorded from mutant EAG1 channels lacking the N-terminal PAS domain (ΔPAS) in the absence and presence of imipramine. The deletion of the PAS domain substantially increased the IC50 of the current inhibition by imipramine and also decreased the efficacy (maximal inhibition) observed at the highest examined ligand concentrations, suggesting that direct binding of imipramine to the PAS domain is contributing to the EAG1 current inhibition (Fig. 2, B and C, and Table 1). The residual inhibition by imipramine of EAG1 channels lacking the PAS domain is most likely due to the secondary inhibitory mechanism via the pore block, as proposed before (27García-Ferreiro R.E. Kerschensteiner D. Major F. Monje F. Stühmer W. Pardo L.A. Mechanism of block of hEag1 K+ channels by imipramine and astemizole.J. Gen. Physiol. 2004; 124: 301-317Crossref PubMed Scopus (115) Google Scholar).Table 1IC50 (μM) for imipramine inhibition at +50 mV for WT and indicated mutant EAG1 channelsChannelIC50 (μM)p ValueWT58.1 ± 9.7 (5)ΔPAS108.6 ± 12.6 (5)ap < 0.05 by Student's t test. p < 0.05 was considered statistically significant. p-values represent significance of Student's t tests used to compare the IC50 for imipramine inhibition at +50 mV for WT and indicated mutant channels. The number of averaged recordings from different oocytes is indicated in parentheses. The difference between IC50 for ΔPAS and R84G mutants was not statistically significant (p value of 0.2067). The difference between IC50 for ΔPAS and D39G/R84G mutants was not statistically significant (p value of 0.0718).0.0131Y71G26.7 ± 3.2 (4)ap < 0.05 by Student's t test. p < 0.05 was considered statistically significant. p-values represent significance of Student's t tests used to compare the IC50 for imipramine inhibition at +50 mV for WT and indicated mutant channels. The number of averaged recordings from different oocytes is indicated in parentheses. The difference between IC50 for ΔPAS and R84G mutants was not statistically significant (p value of 0.2067). The difference between IC50 for ΔPAS and D39G/R84G mutants was not statistically significant (p value of 0.0718).0.0279Y71V31.3 ± 1.8 (6)ap < 0.05 by Student's t test. p < 0.05 was considered statistically significant. p-values represent significance of Student's t tests used to compare the IC50 for imipramine inhibition at +50 mV for WT and indicated mutant channels. The number of averaged recordings from different oocytes is indicated in parentheses. The difference between IC50 for ΔPAS and R84G mutants was not statistically significant (p value of 0.2067). The difference between IC50 for ΔPAS and D39G/R84G mutants was not statistically significant (p value of 0.0718).0.0153Y71F58.0 ± 3.7 (6)0.9920Y71E28.5 ± 1.4 (5)ap < 0.05 by Student's t test. p < 0.05 was considered statistically significant. p-values represent significance of Student's t tests used to compare the IC50 for imipramine inhibition at +50 mV for WT and indicated mutant channels. The number of averaged recordings from different oocytes is indicated in parentheses. The difference between IC50 for ΔPAS and R84G mutants was not statistically significant (p value of 0.2067). The difference between IC50 for ΔPAS and D39G/R84G mutants was not statistically significant (p value of 0.0718).0.0166Y71R43.4 ± 7.5 (4)0.2892D39G79.6 ± 5.4 (6)0.0731V80G57.1 ± 3.1 (6)0.9176V83G44.3 ± 2.6 (6)0.1685R84G92.6 ± 2.7 (6)ap < 0.05 by Student's t test. p < 0.05 was considered statistically significant. p-values represent significance of Student's t tests used to compare the IC50 for imipramine inhibition at +50 mV for WT and indicated mutant channels. The number of averaged recordings from different oocytes is indicated in parentheses. The difference between IC50 for ΔPAS and R84G mutants was not statistically significant (p value of 0.2067). The difference between IC50 for ΔPAS and D39G/R84G mutants was not statistically significant (p value of 0.0718).0.0047F87G42.8 ± 0.6 (5)0.1541F130G52.8 ± 5.0 (6)0.6211D39G/R84G142 ± 28.5 (6)ap < 0.05 by Student's t test. p < 0.05 was considered statistically significant. p-values represent significance of Student's t tests used to compare the IC50 for imipramine inhibition at +50 mV for WT and indicated mutant channels. The number of averaged recordings from different oocytes is indicated in parentheses. The difference between IC50 for ΔPAS and R84G mutants was not statistically significant (p value of 0.2067). The difference between IC50 for ΔPAS and D39G/R84G mutants was not statistically significant (p value of 0.0718).0.0060a p < 0.05 by Student's t test. p < 0.05 was considered statistically significant. p-values represent significance of Student's t tests used to compare the IC50 for imipramine inhibition at +50 mV for WT and indicated mutant channels. The number of averaged recordings from different oocytes is indicated in parentheses. The difference between IC50 for ΔPAS and R84G mutants was not statistically significant (p value of 0.2067). The difference between IC50 for ΔPAS and D39G/R84G mutants was not statistically significant (p value of 0.0718). Open table in a new tab To determine the structural basis of imipramine binding to the PAS domain, we performed computational modeling. While KCNH are the only known ion channels to include PAS domains as part of their amino acid sequence, PAS domain fold is quite common in non-ion channel proteins. In the PAS domains of other proteins, ligands typically bind inside the PAS domain cavity formed by the antiparallel β-strands sandwiched between the α-helices (28Henry J.T. Crosson S. Ligand-binding PAS domains in a genomic, cellular, and structural context.Annu. Rev. Microbiol. 2011; 65: 261-286Crossref PubMed Scopus (288) Google Scholar, 29Möglich A. Ayers R.A. Moffat K. Structure and signaling mechanism of per-ARNT-sim domains.Structure. 2009; 17: 1282-1294Abstract Full Text Full Text PDF PubMed Scopus (401) Google Scholar, 30Key J. Hefti M. Purcell E.B. Moffat K. Structure of the redox sensor domain of Azotobacter vinelandii NifL at atomic resolution: signaling, dimerization, and mechanism.Biochemistry. 2007; 46: 3614-3623Crossref PubMed Scopus (93) Google Scholar). However, the initial docking of imipramine into the PAS domain cavity using AutoDock Vina (31Trott O. Olson A.J. AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading.J. Comput. Chem. 2010; 31: 455-461Crossref PubMed Scopus (7925) Google Scholar) indicated that Y71 blocks the entrance to the cavity for imipramine (shown in grey in Fig. 2, D). To sample thermodynamically available conformations, we used replica exchange solute tempering (REST2) (32Liu P. Kim B. Friesner R.A. Berne B.J. Replica exchange with solute tempering: a method for sampling biological systems in explicit water.Proc. Natl. Acad. Sci. U. S. A. 2005; 102: 13749-13754Crossref PubMed Scopus (584) Google Scholar) to elucidate the ligand-accessible PAS domain cavity. The REST2 simulations indicated that in the ligand-accessible conformation, Y71 swings away from the binding pocket, no longer blocking imipramine binding inside the cavity (shown in green in Fig. 2D). Consistent with these results, replacing the tyrosine at position 71 with a smaller glycine and valine increased EAG1 current inhibition for imipramine (Figs. 2, B, E and F and S1, A and C, and Table 1), while replacing the tyrosine with structurally similar phenylalanine had no statistically significant effect on imipramine inhibition (Fig. S1, B and D). These results strongly suggested that Y71 functions as a "gate-keeper" residue, restricting access of imipramine to the PAS domain cavity. To further investigate if the charge at position 71 may have an effect on EAG1 current inhibition by imipramine, Y71 was substituted for either glutamate or arginine. Imipramine inhibition of EAG channels with Y71E or Y71R mutation was not statistically significantly different from each other, suggesting that the effect of Y71 on EAG1 current inhibition by imipramine does not involve electrostatic interactions. The analysis of the electrostatic profile of the PAS domain indicated that the PAS domain cavity is lined with predominantly hydrophobic residues (Fig. 3A). The hydrophobic profile of the cavity should facilitate the binding of the hydrophobic tricyclic ring structure of imipramine. To investigate the dynamics of imipramine binding, the initial poses for imipramine binding obtained with the REST2 simulations were subjected to 100 ns MD simulations. The MD simulations indicated that imipramine remained in the binding pocket with its three rings facing the inside of the PAS domain cavity (Fig. 3B). To further investigate the contribution of the hydrophobic profile of the cavity we in silico generated mutant PAS domains with the cavity lining hydrophobic residues V80, V83, F87 and F130 mutated to glutamates (4E mutant) or phenylalanines (4T mutant). The mutant PAS domain structures were generated using Amber force field with the initial structures taken from the REST2 simulations. Imipramine was docked into the mutant PAS domains and subjected to 100 ns MD simulations. For the 4E mutant PAS domain imipramine drifted away from the cavity after the first 10 ns (Fig. 3C), while for the 4T mutant PAS domain imipramine stayed inside the cavity for the duration of the simulation (Fig. 3D). We next calculated imipramine binding free energies for the wild-type (WT) and two mutant PAS domains using MM-PBSA (33Wang C. Greene D. Xiao L. Qi R. Luo R. Recent developments and applications of the MMPBSA method.Front. Mol. Biosci. 2018; 4: 87Crossref PubMed Scopus (326) Google Scholar). Different components of the binding free energy, including van der Waals (VDW), electrostatic, polar, and non-polar solvation energies, are given in Table 2. The breakdown of the binding free energy indicates that the binding is driven by hydrophobic interactions between the hydrophobic residues in the cavity and the hydrophobic rings of imipramine. Consistent with this, the imipramine binding free energy for the 4E mutant PAS domain was significantly reduced, while for the 4T mutant, it was the same as for the WT PAS domain (Table 2). Taken together, the results of the MD simulations for the WT and two mutant PAS domains suggest that the hydrophobic profile of the cavity facilitates imipramine binding.Table 2Imipramine binding free energies (in kcal/mol) calculated with MM-PBSAPAS domainVDWElecPolar solvNon-polar solvTotalWT−42.72 ± 0.67−0.20 ± 0.112.59 ± 0.17−4.41 ± 0.02−34.74 ± 0.254E−23.40 ± 0.31−0.81 ± 0.078.25 ± 0.23−2.78 ± 0.02−18.74 ± 0.314T−41.96 ± 0.30−0.80 ± 0.0411.96 ± 0.11−4.18 ± 0.01−34.98 ± 0.28Abbreviations: elec, electrostatic; non-polar solv, non-polar solvation; polar solv, polar solvation; VDW, Van der Waals.Total energies were computed with MM-PBSA. Open table in a new tab Abbreviations: elec, electrostatic; non-polar solv, non-polar solvation; polar solv, polar solvation; VDW, Van der Waals. Total energies were computed with MM-PBSA. The computational modeling suggested that residues D39, V80, V83, R84, F87, and F130 in the PAS domain cavity may be contributing to the coordination of imipramine inside the PAS cavity (Fig. 4A). To investigate the contribution of these residues to the imipramine inhibition of EAG1 currents, they were substituted one by one for glycine, and currents from the mutant channels were recorded over the range of imipramine concentrations (Fig. 4, B–H). Substituting glycine for R84 showed a statistically significant increase in the IC50 of EAG1 current inhibition and this increase was statistically the same as the one observed for the ΔPAS mutation (Fig. 4I and Table 1). The other five mutations did not have a substantial effect on the IC50 of imipramine inhibition (Fig. 4I and Table 1). Interactions mediated by van der Waals forces and potential π−π stacking are energetically weaker than electrostatic interactions. Therefore, the absence of an effect on the IC50 for the individual valine and phenylalanine substitutions is not surprising. MD simulations suggest that these hydrophobic residues are collectively contributing to the ligand coordination and binding. To experimentally test this possibility, we introduced double mutations F87G/F130G, V80G/F87G, and F87G/F130G. Unfortunately, these double mutant channels did not generate any detectable currents. Therefore, we were unable to experimentally determine the collective contribution of the hydrophobic residues to the imipramine binding and coordination in the PAS domain cavity. Although the D39G mutation did not change the IC50 in a statistically significant manner, it caused a similar decrease in the current inhibition as R84G mutation at high imipramine concentrations (Fig. 4I). To further examine the contribution of D39 and R84 residues to the current inhibition, we generated a double mutant D39G/R84G EAG1 channels. The double mutant channels generated currents and showed an even larger decrease in the current inhibition at 300 μM imipramine concentration than the one observed for the individual mutations (Fig. 4, H and I, and Table 1). Moreover, the decrease in the current inhibition was comparable to the one observed for ΔPAS mutant channels. Taken together, these results suggest that D39G and R84G residues are essential for EAG1 current inhibition by imipramine binding to the PAS domain. It is possible that the considered mutations in EAG1 channels substantially alter channel gating obscuring the interpretation of the mechanisms of imipramine inhibition. Contrary to this possibility, the current profile in the absence of the inhibitors was very similar for all mutant channels tested, except ΔPAS mutant that had been shown to affect the voltage-dependence of EAG1 channels (34Carlson A.E. Brelidze T.I. Zagotta W.N. Flavonoid regulation of EAG1 channels.J. Gen. Physiol. 2013; 141: 347-358Crossref PubMed Scopus (30) Google Scholar, 35Terlau H. Heinemann S.H. Stühmer W. Pongs O. Ludwig J. Amino terminal-dependent gating of the potassium channel rat eag is compensated by a mutation in the S4 segment.J. Physiol. 199

Referência(s)