Structural characterization of voltage-gating and ion-permeation mechanisms in connexin hemichannels

Abstract

This study investigated the molecular mechanisms underlying the voltage gating and the selective ion-permeation of connexin hemichannels through characterizing atomic structures. Two structural works of connexin channels guided this study: 1) the solution of 'open' Cx26 gap junction channel structure in 3.5A resolution by the X-ray crystallography (Maeda et al., 2009) and 2) the identification of the permeability barrier of loop-gate 'closed' state of Cx32*43E1 hemichannel by metal-bridge crosslinking experiments (Tang et al., 2009). I further characterized the open and the closed hemichannel structures.;Comparison of experimental current-voltage (I-V) relations of open Cx26 hemichannels with those computed with GCMC/BD simulations indicated that the Cx26 crystal structure refined by MD simulations and known co- and post-translational modifications corresponds to the physiological open state. The GCMC/BD simulations of the refined open structure revealed the molecular mechanisms underlying charge-selective ion-permeation.;Because voltage is known to initiate connexin hemichannel closure by destabilization of the open state, the MD simulations of the open Cx26 hemichannel were further analyzed. The simulations identified large interconnected van der Waals and electrostatic networks that stabilize the region of the channel pore forms a permeability barrier in the loop-gating process. Also, dynamic fluctuations in the electrostatic network were directly correlated with the parameters that define the structure of the parahelix. This suggests that the electrostatic network forms at least a portion of the loop-gate voltage-sensor and that the voltage-sensor is directly coupled to the gate.;Currently, there is no atomic resolution structure of a connexin-closed state. A powerful alternative method to define the structure is a biochemical method using metal-bridge formation with substituted cysteines. The closed structures of the other regions were characterized including the intracellular and extracellular entrances of their pore by using the same method. L108C and E109C at the intracellular entrance had reversible metal-bridge formation. However, no cysteine mutants at the extracellular entrance exhibited metal-bridge formation. These results indicated that only the intracellular entrance of the loop-gated closed structure has the pore size reduced from ∼15A to ∼10A in diameter. The distance constraints provide a means to build the atomic model of the loop-gate closed state.