Channelrhodopsin C1C2: Photocycle kinetics and interactions near the central gate

Abstract

Channelrhodopsins (ChR) are light-sensitive cation channels used in optogenetics, a technique that applies light to control cells (e.g., neurons) that have been modified genetically to express those channels. Although mutations are known to affect pore kinetics, little is known about how mutations induce changes at the molecular scale. To address this issue, we first measured channel opening and closing rates of a ChR chimera (C1C2) and selected variants (N297D, N297V, and V125L). Then, we used atomistic simulations to correlate those rates with changes in pore structure, hydration, and chemical interactions among key gating residues of C1C2 in both closed and open states. Overall, the experimental results show that C1C2 and its mutants do not behave like ChR2 or its analogous variants, except V125L, making C1C2 a unique channel. Our atomistic simulations confirmed that opening of the channel and initial hydration of the gating regions between helices I, II, III, and VII of the channel occurs with 1) the presence of 13-cis retinal; 2) deprotonation of a glutamic acid gating residue, E129; and 3) subsequent weakening of the central gate hydrogen bond between the same glutamic acid E129 and asparagine N297 in the central region of the pore. Also, an aspartate (D292) is the unambiguous primary proton acceptor for the retinal Schiff base in the hydrated channel.

Figure 1

(A) Simulation snapshot of closed-state wild type channelrhodopsin chimera (C1C2-WT). (B) Four-state ChR2 photocycle model, with two closed (C1 and C2) and two open (O1 and O2) states. G_d1 and G_d2 are the rates of O1 → C1 and O2 → C2 transitions, respectively. G_r is the rate of thermal C2 → C1 conversion and recovery of the initial closed state. The rates of transitions between open states 1 and 2 are e₁₂ and e₂₁. Activation rates k₁ for C1 → O1 and k₂ for C2 → O2 processes are associated with protein conformational changes related to absorption of a photon of light. (C) Simulation snapshot of open-state C1C2-WT. Water, represented as blue surfaces, fills the open channel but not the closed channel.

Figure 2

(A) Location of mutated residues V125 (blue) and N297 (yellow) in the transmembrane region of equilibrated, closed-state C1C2-WT (gray cylinders). Key residues involved in protein gating are shown as sticks. For clarity, water, lipids, and ions are not shown. Simulation snapshot of equilibrated (B) closed-state and (C) open-state C1C2-WT seen from the extracellular side of the membrane. Water molecules present within 5 Å of the central gating residues, E129(90) and N297(258), are shown as blue surfaces. Residue locations are as follows: V125 and E129 (helix II); E162 and C167 (helix III); D195 (helix IV); and N297, D292, and retinal (helix VII). Retinal is attached to helix VII through a covalent bond with K296 (helix VII), where a Schiff base forms.

Figure 3

Kinetic analysis of photocurrents and reversal potential determination from two-electrode voltage clamp recordings. Light activation is indicated by the blue bars. (A) Indication of characteristic parts of the photocurrent curve used to analyze channel kinetics. (B) Normalized photocurrent traces of C1C2 constructs expressed in Xenopus oocytes, V_m = −100 mV, Na⁺ solution (pH 7.0). Significant changes to photocycle kinetics were seen for all mutants, with V125L showing the greatest change overall. Individual rate values and significance of changes to each rate are reported in Table 1. The colors represent the following: blue, WT; magenta, V125L; red, N297D; and green, N297V. (C) Photocurrent traces for C1C2 wild-type and single mutant constructs recorded using the inset voltage protocol in Na⁺ measuring solution (pH 7.0). Scale bar is for each representative trace and current-voltage relationship for I_p (blue) and I_ss (orange) representing average ± standard error of the mean. Reversal potentials were determined from current-voltage plots at the V_m where the current is zero.