Breaking the hydrophobicity of the MscL pore: insights into a charge-induced gating mechanism

Abstract

The mechanosensitive channel of large conductance (MscL) is a protein that responds to membrane tension by opening a transient pore during osmotic downshock. Due to its large pore size and functional reconstitution into lipid membranes, MscL has been proposed as a promising artificial nanovalve suitable for biotechnological applications. For example, site-specific mutations and tailored chemical modifications have shown how MscL channel gating can be triggered in the absence of tension by introducing charged residues at the hydrophobic pore level. Recently, engineered MscL proteins responsive to stimuli like pH or light have been reported. Inspired by experiments, we present a thorough computational study aiming at describing, with atomistic detail, the artificial gating mechanism and the molecular transport properties of a light-actuated bacterial MscL channel, in which a charge-induced gating mechanism has been enabled through the selective cleavage of photo-sensitive alkylating agents. Properties such as structural transitions, pore dimension, ion flux and selectivity have been carefully analyzed. Besides, the effects of charge on alternative sites of the channel with respect to those already reported have been addressed. Overall, our results provide useful molecular insights into the structural events accompanying the engineered MscL channel gating and the interplay of electrostatic effects, channel opening and permeation properties. In addition, we describe how the experimentally observed ionic current in a single-subunit charged MscL mutant is obtained through a hydrophobicity breaking mechanism involving an asymmetric inter-subunit motion.

Fig 1. Structure of Tb-MscL and photo-sensitive compound.

(A) X-ray crystallographic structure of Tb-MscL (PDB ID: 2OAR) with the transmembrane helices TM1 and TM2 colored in yellow and blue, respectively. Locations of residues lining the inner helix (TM1) are indicated. (B) Single subunit of Tb-MscL. (C) Structure of photo-sensitive compound that releases the charged group upon light irradiation. (D) Schematic view of the step-by-step substitution process followed in the MD simulation protocol.

Fig 2. Average pore radius along the channel.

Average pore radius along the channel axial position (Z-coordinate) with respect to the mutated site (shaded portion, Z = 0 Ang). Inset plot represents the distribution of minimum pore radius calculated by limiting the analysis to the constricted zone of the pore (residues 17 to 23).

Fig 3. PMF profile, water and ion coordination around the potassium ion.

(A) PMF profile as a function of the z-coordinate. The origin was set at the C^α geometric centroid of the mutated site. Vertical dashed lines indicate the average location of the acetate group oxygens centroid with respect to the origin. (B) Schematic view of ion exit pathway from the mutated site. The charged group (NL) and the bulky ligand (1L) are shown in yellow. Blue spheres inside the pore represents the K⁺ ion. (C) Average number of water molecules and (D) potassium ions within a radius of 3.5 and 5 Ang, respectively, around the permeating K⁺ ion. Data obtained with a smaller radius (4 Ang) is shown in the inset.

Fig 4. Ion occupancy and histogram of ion counts along the axial positions.

Occupancy of K⁺ ion inside the pore obtained by extracting the ions within a distance of 10 Ang from the mutated site along the radial and axial dimensions. Right panel shows the histogram of K⁺ counts along the axial positions for NL (black) and 1L (blue) models.

Fig 5. Average pore radius and protein-water interaction at cytoplasmic entrance.

(A) Average pore radius along the axial position. (B) Representative snapshot of the 17M model with charged groups shown in green. (C) Interaction of charged groups with water molecules within a distance of 3 Ang (17M model). Protein-water hydrogen bonds are shown as blue springs.

Fig 6. Pore radius, hydration along the channel and intersubunit contacts.

Pore radius along the axial positions as a function of simulation time for (A) WT, (B) WT_1e and (C) NL models, obtained considering the backbone atoms. (D) Distribution of water molecules in the pore. For clarity, water molecules at the periplasmic vestibule of the pore are shown in black. (E) Inter-helix (TM1 vs TM1) contacts in WT_1e simulation relative to the starting configuration as a function of time.