MerMAIDs: a family of metagenomically discovered marine anion-conducting and intensely desensitizing channelrhodopsins

Abstract

Channelrhodopsins (ChRs) are algal light-gated ion channels widely used as optogenetic tools for manipulating neuronal activity. ChRs desensitize under continuous bright-light illumination, resulting in a significant decline of photocurrents. Here we describe a metagenomically identified family of phylogenetically distinct anion-conducting ChRs (designated MerMAIDs). MerMAIDs almost completely desensitize during continuous illumination due to accumulation of a late non-conducting photointermediate that disrupts the ion permeation pathway. MerMAID desensitization can be fully explained by a single photocycle in which a long-lived desensitized state follows the short-lived conducting state. A conserved cysteine is the critical factor in desensitization, as its mutation results in recovery of large stationary photocurrents. The rapid desensitization of MerMAIDs enables their use as optogenetic silencers for transient suppression of individual action potentials without affecting subsequent spiking during continuous illumination. Our results could facilitate the development of optogenetic tools from metagenomic databases and enhance general understanding of ChR function.

Fig. 1

Discovery and electrophysiological features of MerMAID.s a Unrooted phylogenetic tree of the channelrhodopsin superfamily, with gray circles representing bootstrap values >90%. Scale bar indicates the average number of amino acid substitutions per site. CCR, cation-conducting channelrhodopsin; ACR, anion-conducting channelrhodopsin. An overview of ChRs used to generate the phylogenetic tree can be found in the Supplementary Data 1. b Distribution and relative abundance of MerMAIDs in samples from the Tara Oceans project. Area of each circle indicates the estimated average abundance of MerMAID-like rhodopsins at different Tara Oceans stations. Stations were MerMAIDs were not detected (n.d.) are indicated by crosses. c Photocurrent traces of representative members of previously identified ChR families and MerMAIDs, recorded from −60 to +40 mV in steps of 20 or 15 mV (GtCCR4). Gray bars indicate light application at denoted wavelengths. d, e Desensitization (d) and peak current amplitudes (e) of all MerMAIDs at −60 mV during continuous illumination with 500 nm light. f Normalized action spectrum of MerMAID1. Single measurements are shown as dots (n = 4), and the solid line represents fitted data. Dashed lines indicate light penetration depth in coastal and open seawater (adopted from ref. ). λ_max, maximum response wavelength; g λ_max for all MerMAIDs. Mean values (thick lines) ± standard deviation (whiskers) are shown, and single-measurement data points are represented as dots. Source data are provided as a Source Data file (d–g)

Fig. 2

Ion selectivity and kinetic properties of MerMAIDs a Representative photocurrent traces of MerMAID1 elicited with 500 nm light (gray bar) at different membrane potentials (−80 to +40 mV, in 20 mV steps, from bottom to top) before (left, gray) and after extracellular chloride reduction (right, cyan), as indicated. Insets show enlarged views of the remaining stationary photocurrent. b Current-voltage relationship of the MerMAID1 peak photocurrent at 150 mM (gray) and 10 mM (cyan) extracellular chloride ([Cl⁻]_e). Arrows indicate reversal potentials (E_rev). c Reversal potential shifts (ΔE_rev) upon reduction of [Cl⁻]_e for peak currents of all MerMAIDs as well as the stationary current of MerMAID1. d ΔE_rev values of MerMAID1 upon exchange of external buffer. ΔE_rev of the theoretical Nernst potential for Cl⁻ is indicated as a dashed line (c, d). e Extracellular pH (pH_e) dependence of biphasic MerMAID1 desensitization kinetic. Inset shows the time constants and their relative amplitudes to total decay. f pH_e dependency of the apparent desensitization time constant (τ_des) at −80 and +40 mV. g Voltage dependency of τ_des for all MerMAIDs in ms/mV. h Double-light pulse experiment at −60 mV and pH_e 7.2 to determine the peak current recovery time constant (τ_rec). i pH_e dependency of MerMAID1 at −60 mV. j Recovery time constants of all MerMAID variants. Mean values (thick lines) ± standard deviation (whiskers) are shown, and single-measurement data points are represented as dots. Source data are provided as a Source Data file (b–d, e–g, i, and j)

Fig. 3

Spectroscopic characterization of purified MerMAID1. a Normalized UV/vis absorption spectra of dark-adapted and illuminated MerMAID1. Filled circles indicate single-measurement action spectra recordings, as shown in Fig. 1f. b Normalized UV/vis absorption spectra of MerMAID1 at different pH values, titrated from pH 7.8 to 10.4. The pK values for specific wavelengths are indicated. c Transient absorption changes and electrophysiological recordings obtained with single-turnover laser pulse excitation. d Fine-structured difference absorption spectra obtained from different experiments. (From top to bottom) light minus dark difference spectra obtained from data shown in a, pH-difference spectra calculated from panel b data, evolution-associated difference spectra (EADS) resulting from a global fit of the transient absorption spectra and (bottom) light-minus-dark difference spectrum measured using the FTIR sample shown in g. Due to strong laser scattering, a portion of the spectral data is excluded for the FTIR sample, and residual scattering is marked with an asterisk. e Resonance Raman spectra of dark-adapted MerMAID1 at pH/D 8 (recorded at 488 nm) as well as cryo-trapped and illuminated protein sample at pH 8 (recorded at 413 nm). Inset: zoomed C=NH⁺ stretching region. f Kinetically decomposed FTIR light-minus-dark absorption of MerMAID1, recorded with single turnover and continuous illumination at 0 °C. Bands marked in gray are discussed in the Supplementary Discussion g, Contour plot of transient absorption changes of the sample used in f illuminated with a 532 nm continuous laser. h Kinetics of the fast and slow FTIR components obtained under single-turnover and continuous illumination conditions, respectively. Kinetics at 366 and 500 nm obtained from the UV/vis spectroscopic measurements shown in g are shown for comparison

Fig. 4

MD simulations and mutational analysis of MerMAID1. a Overview of the MD simulation homology model of MerMAID1 in the dark. The predicted ion permeation pathway is shown as mesh (b1, b2), and ribbons represent the protein backbone. b Electrostatic surface potential of the predicted chloride permeation pathway. c Detailed view of the active-site residues, with amino acids shown as cyan sticks and the all-trans retinal (ATR) in orange. Red spheres denote water molecules that remained stable during MD simulation. d Representative photocurrent traces of wild-type (WT) MerMAID1 and selected MerMAID1 mutants recorded at −60 mV. Photocurrent amplitudes (e), λ_max (f), apparent τ_des of the peak current (g), recovery time constant, τ_rec (h), and extent of desensitization (i) of WT MerMAID1 and indicated mutants. Mean values (thick lines) ± standard deviation (whiskers) are shown, and single-measurement data points are represented as dots. Source data are provided as a Source Data file (e–i)

Fig. 5

Neuronal application of MerMAID6 as optogenetic silencer. a CA1 pyramidal neuron expressing MerMAID6-Citrine (green) 5 days after electroporation (stitched maximum intensity projections of two-photon images). mCerulean (magenta) was co-electroporated to visualize neuronal morphology (left). Fluorescence intensity shown as inverted gray values (right). b, c Voltage traces in response to depolarizing current ramps injected into MerMAID6-expressing CA1 pyramidal cells. Illumination with green light (500 nm, 10 mW/mm2) for a brief (10 ms, b) or longer (500 ms, c) time period blocked single spikes. Light onset preceded action potential onset (measured in the dark condition) by 5 ms. d Same as c but a depolarizing current step of 300 pA was injected into the neuron instead of a current ramp