Mapping ligand binding pockets in chloride ClC-1 channels through an integrated in silico and experimental approach using anthracene-9-carboxylic acid and niflumic acid

Abstract

Background and purpose: Although chloride channels are involved in several physiological processes and acquired diseases, the availability of compounds selectively targeting CLC proteins is limited. ClC-1 channels are responsible for sarcolemma repolarization after an action potential in skeletal muscle and have been associated with myotonia congenita and myotonic dystrophy as well as with other muscular physiopathological conditions. To date only a few ClC-1 blockers have been discovered, such as anthracene-9-carboxylic acid (9-AC) and niflumic acid (NFA), whereas no activator exists. The absence of a ClC-1 structure and the limited information regarding the binding pockets in CLC channels hamper the identification of improved modulators.

Experimental approach: Here we provide an in-depth characterization of drug binding pockets in ClC-1 through an integrated in silico and experimental approach. We first searched putative cavities in a homology model of ClC-1 built upon an eukaryotic CLC crystal structure, and then validated in silico data by measuring the blocking ability of 9-AC and NFA on mutant ClC-1 channels expressed in HEK 293 cells.

Key results: We identified four putative binding cavities in ClC-1. 9-AC appears to interact with residues K231, R421 and F484 within the channel pore. We also identified one preferential binding cavity for NFA and propose R421 and F484 as critical residues.

Conclusions and implications: This study represents the first effort to delineate the binding sites of ClC-1. This information is fundamental to discover compounds useful in the treatment of ClC-1-associated dysfunctions and might represent a starting point for specifically targeting other CLC proteins.

Figure 1

Effect of 9‐AC and NFA on ClC‐1 channels expressed in HEK 293 cells. Representative current traces of ClC‐1 channels expressed in HEK 293 cells before and after application of (A) 300 μM 9‐AC and (B) 300 μM NFA. Dashed line indicates zero current. The voltage protocol is reported above the current traces. (C) Dose–response relationship of 9‐AC and NFA block of ClC‐1 steady‐state currents measured at −90 mV. Steady‐state current–voltage relationship of ClC‐1 channels before and after (D) 9‐AC and (E) NFA 300 μM. (F) Time course of 300 μM NFA block and washout. Data are mean ± SEM of n = 5 cells. *P < 0.05.

Figure 2

Pockets identified in the ClC‐1 homology model by the FLAPsite algorithm. The ClC‐1 dimer is depicted as a cartoon, with the pockets identified as red (P1), magenta (P2), green (P3) and yellow (P4) surface representations. The pore cavity identified by CAVER 3.0 is depicted by dots (Chovancova et al., 2012). The C‐terminus of both monomers (from residue 598) is not shown for clarity.

Figure 3

Docking simulations of 9‐AC and NFA on ClC‐1. Top‐scored poses of 9‐AC (green) and NFA (magenta) in the ClC‐1 dimeric structure. (A) Suggested docking site of 9‐AC. The zoomed monomer is depicted as a cartoon, while critical amino acid residues (blue) and 9‐AC (green) are shown as sticks. Dotted lines depict the salt‐bridge interactions between the carboxyl group of 9‐AC and amine and guanidinium groups of K231 and R421, respectively, and the H‐bond interaction between the backbone of R421 and 9‐AC. (B) Suggested docking site of NFA. The monomer is depicted as a cartoon, while critical residues (blue) and NFA (magenta) are shown as sticks. The salt‐bridge interaction between the carboxyl group of NFA and the guanidine moiety of R421 is presented as a dotted line.

Figure 4

Effect of 9‐AC on steady‐state currents of engineered ClC‐1 mutants. (A) Representative current traces of ClC‐1 mutants expressed in HEK 293 cells before and after the application of 300 μM 9‐AC. (B) Steady‐state current–voltage relationship of the channels indicated before and after the application of 300 μM 9‐AC. Data are mean ± SEM of n = 5 cells. *P < 0.05.

Figure 5

Effect of 9‐AC on steady‐state currents of myotonia congenita ClC‐1 mutants. (A) Representative current traces of the ClC‐1 mutants expressed in HEK 293 cells before and after the application of 300 μM 9‐AC. (B) Steady‐state current–voltage relationship of the channels indicated before and after the application of 300 μM 9‐AC. Data are mean ± SEM of n = 5 cells. ^* P < 0.05.

Figure 6

Effect of NFA on steady state currents of engineered ClC‐1 mutants. (A) Representative current traces of the ClC‐1 mutants expressed in HEK 293 cells before and after the application of 300 μM NFA. (B) Steady‐state current–voltage relationship of the channels indicated before and after the application of 300 μM NFA. Data are mean ± SEM of n = 5 cells. ^* P < 0.05.