Genetic engineering of natural light-gated ion channels has proven a powerful way to generate optogenetic tools for a wide variety of applications. In recent years, blue-light activated engineered anion-conducting channelrhodopsins (eACRs) have been developed, improved, and were successfully applied in vivo. We asked whether the approaches used to create eACRs can be transferred to other well-characterized cation-conducting channelrhodopsins (CCRs) to obtain eACRs with a broad spectrum of biophysical properties. We generated 22 variants using two conversion strategies applied to 11 CCRs and screened them for membrane expression, photocurrents and anion selectivity. We obtained two novel eACRs, Phobos and Aurora, with blue- and red-shifted action spectra and photocurrents similar to existing eACRs. Furthermore, step-function mutations greatly enhanced the cellular operational light sensitivity due to a slowed-down photocycle. These bi-stable eACRs can be reversibly toggled between open and closed states with brief light pulses of different wavelengths. All new eACRs reliably inhibited action potential firing in pyramidal CA1 neurons. In Drosophila larvae, eACRs conveyed robust and specific light-dependent inhibition of locomotion and nociception.

Genetic engineering of natural light-gated ion channels has proven a powerful way to generate optogenetic tools for a wide variety of applications. In recent years, blue-light activated engineered anion-conducting channelrhodopsins (eACRs) have been developed, improved, and were successfully applied in vivo. We asked whether the approaches used to create eACRs can be transferred to other well-characterized cation-conducting channelrhodopsins (CCRs) to obtain eACRs with a broad spectrum of biophysical properties. We generated 22 variants using two conversion strategies applied to 11 CCRs and screened them for membrane expression, photocurrents and anion selectivity. We obtained two novel eACRs, Phobos and Aurora, with blue-and red-shifted action spectra and photocurrents similar to existing eACRs. Furthermore, step-function mutations greatly enhanced the cellular operational light sensitivity due to a slowed-down photocycle. These bi-stable eACRs can be reversibly toggled between open and closed states with brief light pulses of different wavelengths. All new eACRs reliably inhibited action potential firing in pyramidal CA1 neurons. In Drosophila larvae, eACRs conveyed robust and specific light-dependent inhibition of locomotion and nociception.The discovery of natural anion-conducting channelrhodopsins (nACRs) 1-3 and the development of engineered anion-conducting channelrhodopsins (eACRs) by targeted mutagenesis of cation-conducting channelrhodopsins (CCRs) 4-7 introduced a new class of optogenetic tools 8 . The existing eACRs were derived from either Chlamydomonas reinhardtii channelrhodopsin-2 (CrChR2) 7 or the channelrhodopsin chimera C1C2 5 using two complementary strategies. Exchange of a single glutamate for an arginine in the central gate of CrChR2 was sufficient to invert selectivity from cations to anions. Additional exchange of two glutamate residues in the outer pore and the inner gate completely eliminated residual proton conductance, yielding the highly anion-selective eACR iChloC 6 . In parallel, mutation of several amino acids within C1C2 to render the electrostatic potential of the conducting pore more positive, strongly favored anion conductance 5 . Further improvements led to a second highly anion-selective eACR iC++ and the related step function version SwiChR++ 4 . These improved versions have been successfully used to silence neurons in mice or rats in vivo 4,6,[9][10][11][12][13] . We asked whether the approaches used to create eACRs can be transferred to other known CCRs to obtain eACRs with a broad spectrum of biophysical properties, especially different kinetics and spectral sensitivities. So far, all eACRs show action spectra similar to CrChR2 with maximal activation in the blue spectral range. ACRs with a red-shifted absorption maximum are desirable for three reasons: First, long-wavelength light penetrates deeper into biological tissue due to lower absorption and scattering 14,15 . This enables silencing of larger volumes

This Article contains errors in the Results section under subheading 'Biophysical characterization in HEK cells'. "Because Chronos exhibits no serine at the homologous position of S63, which is a main constituent of the inner gate in CrChR2 we speculated that a differently arranged inner gate could be responsible for the missing photocurrents. Therefore, we created the double mutant Chronos A80S E107R to reconstitute a serine residue at the putative inner gate while rendering the central gate anion conductive. " should read: "Because Chronos exhibits no serine at the homologous position of S63, which is a main constituent of the central gate in CrChR2 we speculated that a differently arranged central gate could be responsible for the missing photocurrents. Therefore, we created the double mutant Chronos A80S E107R to reconstitute a serine residue at the putative central gate while rendering the central gate anion conductive. "