Molecular dynamics simulations of the Cx26 hemichannel: Evaluation of structural models with Brownian dynamics

Abstract

The recently published crystal structure of the Cx26 gap junction channel provides a unique opportunity for elucidation of the structure of the conductive connexin pore and the molecular determinants of its ion permeation properties (conductance, current–voltage [I-V] relations, and charge selectivity). However, the crystal structure was incomplete, most notably lacking the coordinates of the N-terminal methionine residue, which resides within the pore, and also lacking two cytosolic domains. To allow computational studies for comparison with the known channel properties, we completed the structure. Grand canonical Monte Carlo Brownian dynamics (GCMC/BD) simulations of the completed and the published Cx26 hemichannel crystal structure indicate that the pore is too narrow to permit significant ion flux. The GCMC/BD simulations predict marked inward current rectification and almost perfect anion selectivity, both inconsistent with known channel properties. The completed structure was refined by all-atom molecular dynamics (MD) simulations (220 ns total) in an explicit solvent and POPC membrane system. These MD simulations produced an equilibrated structure with a larger minimal pore diameter, which decreased the height of the permeation barrier formed by the N terminus. GCMC/BD simulations of the MD-equilibrated structure yielded more appropriate single-channel conductance and less anion/cation selectivity. However, the simulations much more closely matched experimentally determined I-V relations when the charge effects of specific co- and posttranslational modifications of Cx26 previously identified by mass spectrometry were incorporated. We conclude that the average equilibrated structure obtained after MD simulations more closely represents the open Cx26 hemichannel structure than does the crystal structure, and that co- and posttranslational modifications of Cx26 hemichannels are likely to play an important physiological role by defining the conductance and ion selectivity of Cx26 channels. Furthermore, the simulations and data suggest that experimentally observed heterogeneity in Cx26 I-V relations can be accounted for by variation in co- and posttranslational modifications.

Mechanism of gating by calcium in connexin hemichannels

“…Nevertheless, the initial events seen in the present simulations indicate that the subsequent rearrangements of the network of electrostatic interactions below D50-K61 are among the steps that enhance the probability of transition to a Ca 2+ -induced closed state. The previously equilibrated Cx26 hemichannel and other connexin structural models also associate these residues with electrostatic interactions in the open conformation (21,22).…”

“…We proposed that this interaction is disrupted with increasing extracellular Ca 2+ concentrations, and that this may constitute the initial reaction for hemichannel closure (17). To gain further insights into the gating mechanism mediated by Ca 2+ disruption of the D50-K61 interaction, we performed molecular dynamics simulations, building on previously well-established equilibrated structures of Cx26 hemichannels (22,23). We observed that K + ions accumulate near the extracellular entrance of the pore where several negatively charged residues are located, including D50 (23).…”

Mutations in Cx30 that are linked to skin disease and non-syndromic hearing loss exhibit several distinct cellular pathologies

“…Cx30 shares 76% sequence identity with Cx26 (Grifa et al, 1999), and although the crystal structure of Cx30 has not yet been resolved, the crystal structure of Cx26 has elucidated that the N-terminal tail lines the gap junction channel pore, creating a funnel that restricts channel selectivity based on molecular size, flexibility, charge and charge distribution (Harris, 2007;Kwon et al, 2011;Maeda et al, 2009). In the nonsensory cells of the inner ear, Cx26 and Cx30 intermix to form heteromeric and heterotypic channels (Ahmad et al, 2003;Forge et al, 2003;Marziano et al, 2003;Yum et al, 2007) and are suggested to form functional hemichannels (Gossman and Zhao, 2008;Zhao et al, 2005).…”

“…Journal of Cell Science (2014Science ( ) 127, 1751Science ( -1764Science ( doi:10.1242 second transmembrane domain also lines the channel pore and is involved in intraconnexin interactions with other domains, including the N-terminus, that dictate specific protein conformation (Kwon et al, 2011;Maeda et al, 2009). By analogy, it is possible that the A88V Cx30 mutation affects protein folding and stability, and alters important interactions, moving the N-terminal domain away from the cytoplasmic entrance of the pore (Kwon et al, 2011;Maeda et al, 2009) to favor abnormal hemichannel activity.…”

“…By analogy, it is possible that the A88V Cx30 mutation affects protein folding and stability, and alters important interactions, moving the N-terminal domain away from the cytoplasmic entrance of the pore (Kwon et al, 2011;Maeda et al, 2009) to favor abnormal hemichannel activity. Therefore, we propose that leaky A88V Cx30 hemichannels contribute to the induction of apoptosis in REKs.…”