Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 5, 14807

An Improved Chloride-Conducting Channelrhodopsin for Light-Induced Inhibition of Neuronal Activity in Vivo


An Improved Chloride-Conducting Channelrhodopsin for Light-Induced Inhibition of Neuronal Activity in Vivo

Jonas Wietek et al. Sci Rep.


Channelrhodopsins are light-gated cation channels that have been widely used for optogenetic stimulation of electrically excitable cells. Replacement of a glutamic acid in the central gate with a positively charged amino acid residue reverses the ion selectivity and produces chloride-conducting ChRs (ChloCs). Expressed in neurons, published ChloCs produced a strong shunting effect but also a small, yet significant depolarization from the resting potential. Depending on the state of the neuron, the net result of illumination might therefore be inhibitory or excitatory with respect to action potential generation. Here we report two additional amino acid substitutions that significantly shift the reversal potential of improved ChloC (iChloC) to the reversal potential of endogenous GABAA receptors. As a result, light-evoked membrane depolarization was strongly reduced and spike initiation after current injection or synaptic stimulation was reliably inhibited in iChloC-transfected neurons in vitro. In the primary visual cortex of anesthetized mice, activation of iChloC suppressed spiking activity evoked by visual stimulation. Due to its high operational light sensitivity, iChloC makes it possible to inhibit neurons in a large volume of brain tissue from a small, point-like light source.


Figure 1
Figure 1. Strategy for improving the chloride selectivity of ChloC.
(A) Position of key negative (blue) and positive charges (red) in wt Channelrhodopsin, slowChloC, and iChloC with respect to the aqueous pore (light blue). (B) Structural model of ChR2-E90R; view from the intracellular side. Residues having a stake in the inner gate are highlighted in blue and cyan (TM: transmembrane helix). The negatively charged E83 carboxylate is located in the center of the inner gate pore. (C) Closer look at the inner gate residues (side view). Hydrogen bonds may form between E82 and R268 as well as between E83 and H134. Atomic distances (black dotted lines) are shown in Å. The retinal (RET) is shown in orange.
Figure 2
Figure 2. Characterization of slowChloC variants in HEK293 cells.
(A) Photocurrent example traces of slowChloC, slowChloC (E101S), slowChloC (E83Q) and slowChloC (E83Q, E101S) (iChloC) in HEK 293 cells at different holding potentials (20 mV steps). (B) I-E curves for the 4 ChloC variants. (C) Photocurrent amplitudes for the 4 ChloC variants measured at a holding potential of 0 mV. (D) Reversal potential in HEK 293 cells (slowChloC, −52.5 ± 1.3 mV; slowChloC (E101S), −59.2 ± 0.3 mV; slowChloC (E83Q), −60.8 ± 1.5 mV; iChloC, −65.6 ± 1.1 mV). Dashed line indicates calculated Nernst potential for Cl (−69.6 mV). n = 9 for slowChloC, n = 8 for slowChloC (E101S), n = 6 for other two variants.
Figure 3
Figure 3. Expression of iChloC in CA1 pyramidal cells in organotypic hippocampal slice culture.
(A) Overview of neuronal morphology 5 days after electroporation (maximum intensity projection of two-photon images). (B) Fluorescence of co-expressed tdimer2 was used to target transfected neurons for electrophysiological recordings (same cells as in (A)). (C) Dodt contrast image of cells shown in (B). (D) Photocurrents in response to a single light pulse (476 nm, 5 ms, 1 mW/mm2) at different holding potentials. Photocurrents reversed at very negative holding potentials and were large at depolarized holding potentials, where the Cl inward driving force was highest (same neuron as in (AC)). The inset shows the onset of the photocurrents at higher temporal resolution. Indicated holding potentials were rounded to full numbers after subtraction of the liquid junction potential (−10.6 mV). Indicated tau value was derived from 9 independent measurements in 9 slice cultures.
Figure 4
Figure 4. Photocurrent reversal potential vs. GABA receptor reversal potential.
(A) Current traces of CA1 pyramidal cell expressing iChloC. Photocurrent reversal (left) is close to the reversal potential of GABAA receptor mediated IPSCs (right), indicating pure Cl conductance. (B) I-E curves for iChloC photocurrents and IPSCs. (C) Current traces of CA1 pyramidal cell expressing slowChloC. The photocurrent (left) reverses more positive than GABAA receptor mediated IPSCs (right), indicating mixed ionic conductance of slowChloC. (D) I-E curves for slowChloC photocurrents and IPSCs. (E) No systematic difference in GABAA reversal between slowChloC and iChloC neurons. Between individual neurons, GABAA reversal and iChloC reversal were highly correlated (r2 = 0.67), indicating that the same ions are conducted. SlowChloC photocurrents were not correlated with GABAA reversal potential (r2 = 0.04). (F) Light stimulation experiments under current clamp conditions (I = 0) show significantly smaller depolarization from resting membrane potential in iChloC expressing neurons (n = 8) compared to slowChloC expressing neurons (n = 10).
Figure 5
Figure 5. Spike suppression by iChloC.
(A) Voltage traces in response to depolarizing current ramps (0–700 pA) injected into an iChloC-expressing CA1 pyramidal cell. The injected current at the time of the first spike was defined as the rheobase. Illumination with blue light activated iChloC. Increasing light intensities shifted the rheobase to higher values. (B) Individual neurons had different spike thresholds in the dark and showed different operational light sensitivity as a function of iChloC expression level. (C) Number of spikes during the ramp as a function of light intensity. At 0.1 mW/mm2, spike number was reduced to 45%, on average. (D) Suppression of depolarization-induced spiking by iChloC activation was highly reproducible. Due to the slow kinetics of iChloC, 5-ms light pulses were sufficient to block spiking for several seconds.
Figure 6
Figure 6. Suppression of synaptically evoked spikes by iChloC.
(A) Schema of targeted cell-attached recordings from transfected CA1 pyramidal cells in an organotypic hippocampal slice culture. Drawing by J.S. Wiegert & T.G. Oertner. (B) Postsynaptic spikes were induced by electrical stimulation of afferents in stratum radiatum (4 μA, 0.5 ms). Upper trace: electrical stimulation, only. Middle trace: a blue light pulse (0.1 mW/mm2, 5 ms) delivered through the water immersion objective prevented spike initiation after synaptic stimulation. Lower trace: a spike was triggered again by electrical stimulation alone. (C) Raster plot of spike timing after synaptic stimulation of iChloC expressing neuron. Light pulse reliably blocked synaptically evoked spikes (trials #11–20). (D) Summary of synaptic stimulation experiments. While slowChloC activation had no significant effect on electrical spike induction (n = 5 slices cultures), iChloC activation blocked spiking in the majority of neurons (n = 7 slice cultures). Spike inhibition was fully reversible in all experiments.
Figure 7
Figure 7. In vivo suppression of sensory evoked spikes by iChloC.
(A) Schematic representation of the experimental setup. Drawing by R. Beltramo and M. Scanziani. (B) Left panel: Epifluorescence image of a mouse brain expressing iChloC (red) in V1. The black ‘x’ indicates the penetration site of the extracellular electrode that recorded the spikes shown in (C). Right panel: Epifluorescence image of coronal section through V1 of the brain shown on the left. Red: Viral expression of iChloC; Green: Recording site. (C) Extracellular recording from V1 in response to visual stimulation (full-field drifting gratings) under control conditions (black raster plot of multi-unit activity, left panel) and following 100 ms of iChloC photoactivation (blue raster plot, central panel). Control and photostimulation trials were interleaved. Right panel: Average peri-stimulus time histograms (PSTH, binning 50 ms) of control (black) and photostimulation (blue) trials. Striped horizontal bar indicates visual stimulus duration (1.5 s). Vertical cyan line: iChloC photoactivation (100 ms). (D) Same as in (C) for iChloC photostimulation of 10 ms and 5 ms durations. (E) Baseline normalized average PSTH (solid lines) and standard errors (shaded areas) of cortical responses to drifting gratings with (blue) and without (black) iChloC photoactivation (10 ms), from 4 tetrodes in 2 animals.

Similar articles

See all similar articles

Cited by 35 PubMed Central articles

See all "Cited by" articles


    1. Lüthi A. & Lüscher C. Pathological circuit function underlying addiction and anxiety disorders. Nature Neuroscience 17, 1635–1643 (2014). - PubMed
    1. Roux L., Stark E., Sjulson L. & Buzsáki G. In vivo optogenetic identification and manipulation of GABAergic interneuron subtypes. Current Opinion in Neurobiology 26, 88–95 (2014). - PMC - PubMed
    1. Häusser M. Optogenetics: the age of light. Nature Methods 11, 1012–1014 (2014). - PubMed
    1. Liu X., Ramirez S. & Tonegawa S. Inception of a false memory by optogenetic manipulation of a hippocampal memory engram. Philosophical transactions of the Royal Society of London. Series B, Biological sciences 369, 20130142 (2014). - PMC - PubMed
    1. Zhang F. et al. Multimodal fast optical interrogation of neural circuitry. Nature 446, 633–639 (2007). - PubMed

Publication types