Volume 23, Issue 6, Pages 995-1004 (June 2015) Conformational Changes Underlying Desensitization of the Pentameric Ligand-Gated Ion Channel ELIC  Monica N. Kinde, Qiang Chen, Matthew J. Lawless, David D. Mowrey, Jiawei Xu, Sunil Saxena, Yan Xu, Pei Tang  Structure  Volume 23, Issue 6, Pages 995-1004 (June 2015) DOI: 10.1016/j.str.2015.03.017 Copyright © 2015 Elsevier Ltd Terms and Conditions

Structure 2015 23, 995-1004DOI: (10.1016/j.str.2015.03.017) Copyright © 2015 Elsevier Ltd Terms and Conditions

Figure 1 ELIC Cysteine Mutants Retain Functional Properties (A) Side view of the ELIC structure in a closed-channel conformation (PDB 3RQW). For clarity, the cysteine mutation sites are shown in only one subunit, in which the transmembrane (TM) helices 1–4 are marked. The agonist propylamine (PA) binding site in the extracellular domain is highlighted in violet color. Conventional prime numbers (Miller, 1989) for residues lining the pore and in the TM2-TM3 loop are provided along with the sequence numbers. Pore-lining residues (−2′ to 16′) are colored differently for hydrophobic (yellow) and hydrophilic (green) residues. Residues are colored orange in the TM2-TM3 loop, cyan in the TM4, and magenta in the TM3-TM4 loop. (B) Agonist PA-response curves for individual ELIC constructs: wild-type (WT) ELIC (●), cysteine-free ELIC C300A/C313S (○), S229C (■), Q233C (♢), T237C (♦), L240C (), A244C (▽), F247C (□), L253C (), L256C (▴), T260C (△), and A288C (▾). Representative current traces are shown as insets. Vertical and horizontal scale bars indicate 0.2 μA and 1 min, respectively. Data are fit to the Hill equation (see Table S1 for fit information) and reported as the mean ± SEM from n ≥ 5 oocytes. Representative traces showing activation of ELIC with TET labeling are provided in Figure S1. (C) Representative current traces showing desensitization upon prolonged exposure to the agonist PA. Current is expressed as a fraction of maximal current. Structure 2015 23, 995-1004DOI: (10.1016/j.str.2015.03.017) Copyright © 2015 Elsevier Ltd Terms and Conditions

Figure 2 ELIC Conformations in the Resting State (A) 19F-NMR spectra for individual residue positions. The Lorentzian peak fittings for different peak components are shown as gray lines. (B) 19F-NMR linewidths of pore-lining residues (gray bars), obtained from the Lorentzian peak fittings in (A), correlate with backbone pore radii of the ELIC closed-channel structure. The backbone pore radius profile was generated from the ELIC crystal structure (PDB 3RQW). See also Table S2 for peak fitting analysis and Figure S2 demonstrating the pentameric assembly of ELIC both before and after NMR data acquisition. Structure 2015 23, 995-1004DOI: (10.1016/j.str.2015.03.017) Copyright © 2015 Elsevier Ltd Terms and Conditions

Figure 3 Solvent Exposure and Motional Freedom (A) A snapshot of MD simulations of ELIC showing that C313 has greater exposure to water than C300. Red dots are H2O oxygens. For clarity, only the TMD of one subunit is highlighted. (B) The major conformations (darker color) of residues 229 (−2′) and 247 (16′) have greater 19F-NMR linewidths (solid) and smaller temperature coefficients (hash) than the corresponding minor conformations (lighter color). More details of the related NMR spectra and data analysis are provided in Figure S3. (C) A site view of the pore-lining TM2 helices showing major (blue) and minor (green) side-chain conformations of residue 247 (16′) sampled over 50-ns MD simulations. The side chain is attached with the 19F probe and its orientation is defined by the Φ angle between the side chain and membrane normal. (D) A histogram of the Φ angles showing major (blue) and minor (green) orientations of the 247 (16′) side chain. The histogram was generated using a bin size of 1° from 400 snapshots sampled evenly over the MD simulations. (E) The major conformations (blue) have less water molecules within 3 Å of residue 247 (16′) than the minor conformations (green). Structure 2015 23, 995-1004DOI: (10.1016/j.str.2015.03.017) Copyright © 2015 Elsevier Ltd Terms and Conditions

Figure 4 19F-NMR Spectral Comparisons for ELIC in the Resting and Desensitized States APO and PA represent the samples containing no and 18 mM PA agonist, respectively. (A) 19F-NMR spectra of residues in the TM3-TM4 loop (288) and the TM4 helix (300, 313). (B) Spectra of individual residues in the TM2-TM3 loop. (C) Spectra of individual pore-lining residues in the TM2 helix. The Lorentzian peak fittings for different peak components are shown as gray lines. Dashed lines highlight chemical shift changes of residues 237 (6′) and 244 (13′) between the resting and desensitized states. See also Tables S2 and S3 for peak fitting analysis. Structure 2015 23, 995-1004DOI: (10.1016/j.str.2015.03.017) Copyright © 2015 Elsevier Ltd Terms and Conditions

Figure 5 DEER ESR Spectroscopy Revealed Desensitization-induced Increase of Conformational Heterogeneity and Intersubunit Distances (A) Background-subtracted dipolar evolution of DEER data (thin gray lines) fitted with the two-Gaussian model for ELIC 253R1 in the resting (thick black line) state and desensitized (thick gray line) state by 18 mM PA. The desensitized data are offset vertically for clarity. (B) The corresponding inter-spin distance distributions with the mean distance for each peak labeled. Note the broader distance distributions and increased mean distances for the adjacent and nonadjacent subunits in the desensitized ELIC. (C) Top view of ELIC highlighting distances for adjacent and nonadjacent subunits presented in (B). Structure 2015 23, 995-1004DOI: (10.1016/j.str.2015.03.017) Copyright © 2015 Elsevier Ltd Terms and Conditions

Figure 6 Desensitization-induced Conformational Changes in the Transmembrane Pore of ELIC (A) A closed-pore conformation of ELIC's TMD in the resting state (PDB 3RQW). Only two subunits are shown for clarity. (B) Changes in the population-weighted 19F-NMR linewidths for pore-lining residues between the resting (APO) and desensitized (PA-bound) states. The largest increase (dark blue) and decrease (dark green) in linewidths indicate the greatest contraction and expansion in the corresponding pore regions, respectively. (C) A desensitized model of ELIC's TMD mapped with the linewidth changes shown in (B). Residues experiencing expansion and contraction are shown in green and blue, respectively. Structure 2015 23, 995-1004DOI: (10.1016/j.str.2015.03.017) Copyright © 2015 Elsevier Ltd Terms and Conditions

Figure 7 ESR Data Support Expansion and Contraction at the Upper and Lower Pore Regions upon Desensitization (A and B) The deuterium peak at ∼2.2 MHz in the three-pulse ESEEM ESR experiments on the MTSL-labeled ELIC at pore positions (A) 247 (16′) and (B) 233 (2′) show higher and lower solvent accessibility, respectively, for the desensitized (gray) than for the resting (black) channels. The narrow feature at ∼2.2 MHz (shown in circles) is diagnostic of solvent accessibility (Carmieli et al., 2006). (C and D) CW ESR on the MTSL-labeled ELIC at positions (C) 247 (16′) and (D) 233 (2′) show increased and decreased MTSL labeling efficiency, respectively, when the labeling was performed under a desensitization condition (gray) compared with the labeling completed in the absence of agonist (black). The CW ESR data are normalized to the protein concentration. Structure 2015 23, 995-1004DOI: (10.1016/j.str.2015.03.017) Copyright © 2015 Elsevier Ltd Terms and Conditions