We investigated the efficacy of optogenetic inhibition at presynaptic terminals using halorhodopsin, archaerhodopsin and chloride-conducting channelrhodopsins. Precisely timed activation of both archaerhodopsin and halorhodpsin at presynaptic terminals attenuated evoked release. However, sustained archaerhodopsin activation was paradoxically associated with increased spontaneous release. Activation of chloride-conducting channelrhodopsins triggered neurotransmitter release upon light onset. Thus, the biophysical properties of presynaptic terminals dictate unique boundary conditions for optogenetic manipulation.
Subscribe to Journal
Get full journal access for 1 year
only $18.75 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
Zhang, F. et al. Nature 446, 633–639 (2007).
Chuong, A.S. et al. Nat. Neurosci. 17, 1123–1129 (2014).
Chow, B.Y. et al. Nature 463, 98–102 (2010).
Berndt, A., Lee, S.Y., Ramakrishnan, C. & Deisseroth, K. Science 344, 420–424 (2014).
Wietek, J. et al. Science 344, 409–412 (2014).
Govorunova, E.G., Sineshchekov, O.A., Janz, R., Liu, X. & Spudich, J.L. Science 349, 647–650 (2015).
Tye, K.M. et al. Nature 471, 358–362 (2011).
Stuber, G.D. et al. Nature 475, 377–380 (2011).
Spellman, T. et al. Nature 522, 309–314 (2015).
Gradinaru, V. et al. Cell 141, 154–165 (2010).
Fletcher, T.L., Cameron, P., De Camilli, P. & Banker, G. J. Neurosci. 11, 1617–1626 (1991).
Fino, E. & Yuste, R. Neuron 69, 1188–1203 (2011).
Gabernet, L., Jadhav, S.P., Feldman, D.E., Carandini, M. & Scanziani, M. Neuron 48, 315–327 (2005).
Chen, Y.H., Wu, M.L. & Fu, W.M. J. Neurosci. 18, 2982–2990 (1998).
Tombaugh, G.C. & Somjen, G.G. J. Neurophysiol. 77, 639–653 (1997).
Turecek, R. & Trussell, L.O. Nature 411, 587–590 (2001).
Pugh, J.R. & Jahr, C.E. J. Neurosci. 31, 565–574 (2011).
Price, G.D. & Trussell, L.O. J. Neurosci. 26, 11432–11436 (2006).
Raimondo, J.V., Kay, L., Ellender, T.J. & Akerman, C.J. Nat. Neurosci. 15, 1102–1104 (2012).
Cosentino, C. et al. Science 348, 707–710 (2015).
Chen, T.W. et al. Nature 499, 295–300 (2013).
Dreosti, E., Odermatt, B., Dorostkar, M.M. & Lagnado, L. Nat. Methods 6, 883–889 (2009).
Fenno, L.E. et al. Nat. Methods 11, 763–772 (2014).
Grimm, D., Kay, M.A. & Kleinschmidt, J.A. Mol. Ther. 7, 839–850 (2003).
Grünwald, D., Shenoy, S.M., Burke, S. & Singer, R.H. Nat. Protoc. 3, 1809–1814 (2008).
Schindelin, J. et al. Nat. Methods 9, 676–682 (2012).
Mattis, J. et al. Nat. Methods 9, 159–172 (2012).
We thank M. Segal (Weizmann Institute of Science) for discussions and reagents. We thank R. Zwang for help with cloning. We thank Y. Ziv, L. Fenno, I. Lampl, R. Paz and the Yizhar laboratory members for comments on the manuscript. We acknowledge support (to O.Y.) from the Human Frontier Science Program, the I-CORE program of the Planning and Budgeting Committee and the Israel Science Foundation (I-CORE grant no. 51/11, ISF grant no. 1351/12), a European Research Council starting grant (ERC-2013-StG OptoNeuromod 337637), the Adelis Foundation, the Lord Sieff of Brimpton Memorial Fund and the Candice Appleton Family Trust. O.Y. is supported by the Gertrude and Philip Nollman Career Development Chair.
The authors declare no competing financial interests.
Integrated supplementary information
Supplementary Figure 1 eArch3.0 and eNpHR3.0 express in soma and neurite membranes of cultured hippocampal neurons.
Cultured hippocampal neurons were transduced with viral vectors encoding eArch3.0-mCherry (a) or eNpHR3.0-mCherry (b) and stained with anti-synapsin1 antibody. Images depict the expression of these mCherry-fused opsins in the somatic membrane (left, arrows) and in the axonal and dendritic membranes (right; indicated by co-labeling with synapsin I, which localizes predominantly to presynaptic terminals).
Supplementary Figure 2 Effects of eArch3.0 and eNpHR3.0 on evoked feedforward inhibition in S1 L4 neurons.
Characterization of feed-forward inhibition recruitment during presynaptic terminal inhibition in acute brain slices using the microbial rhodopsins eArch3.0 and eNpHR3.0 expressed in the whisking thalamus. The bipolar stimulation electrode was placed in the thalamocortical fiber tract. Whole-cell voltage clamp recordings were performed in layer 4 neurons of the S1 barrel cortex. (a) Representative traces of EPSCs and IPSCs evoked by electrical microstimulation (eEPSCs and eIPSCs). Neurons were clamped to –70 mV or 0 mV to isolate synaptic excitation or inhibition, respectively. The inset shows that the eIPSC delay is longer than the eEPSC delay, indicating a di-synaptic response in case of eIPSCs. The timing of the electrical stimulation is indicated by the black horizontal bar. The schematic shows how activation of thalamocortical projections leads to recruitment of local feed-forward inhibition. (b,f) Representative traces of eIPSCs with and without 590 nm illumination of presynaptic terminals of thalamocortical neurons expressing eArch3.0 (b) or eNpHR3.0 (f). Individual traces are overlaid with their mean. Time of electrical stimulation is indicated by black ticks below the traces. (c,d,g,h) Quantification of the eIPSC amplitude before and during brief light application (200 ms, 590 nm, 2 mW mm–2). Activation of both eArch3.0 (n = 17; 40.4% reduction; P = 4.26x10–3) and eNpHR3.0 (n = 33; 65.6% reduction; P = 1.05x10–4) led to a significant reduction in eIPSC amplitude. A significant increase in paired pulse ratio was observed for eNpHR3.0 but not eArch3.0 (P = 0.02 and P = 0.071 for eNpHR3.0 and eArch3.0, respectively). (e,i) eIPSC amplitude relative to baseline during 5 min constant illumination. (e) eArch3.0 activation caused a significant reduction in eIPSC amplitude, with a gradual increase in the efficacy of inhibition of eIPSCs during the first 3 min of light application (n = 10). Inhibition of eIPSCs persisted the entire 5 min after light measurement. (i) eNpHR3.0 activation caused a significant reduction in eIPSC amplitude during constant light delivery (n = 14). Error bars indicate s.e.m. (* P < 0.05; see Supplementary Table 1 for statistics).
Supplementary Figure 3 Rebound responses evoked by light offset can be attenuated by ramp-like light termination.
(a) Voltage clamp recordings from L4 barrel cortex neurons showing EPSCs evoked by electrical stimulation of thalamocortical axons (left). Example of responses that are effectively eliminated (right) by 590 nm illumination of presynaptic terminals in slices expressing eNpHR3.0 in thalamocortical neurons. Individual response traces (light blue) are overlaid with their mean (dark blue). Time of electrical microstimulation is indicated by black ticks above the traces. Note the rebound currents evoked by light offset. (b) Rebound EPSCs evoked by light offset (left) can be attenuated by replacement of step-like termination of the light pulse with a ramp-like reduction in light power (right). Light power during the protocol is depicted by the yellow traces (bottom). (c) Quantification of rebound charge (n = 8). Rebound responses are measured after light offset in recordings in which light is gradually reduced over varying durations (toff). Gradual light termination over 60 ms (33 mW mm–2 s–1) or more did not evoke rebound responses in any of the measured cells. Error bars indicate s.e.m. (see Supplementary Table 1 for statistics).
Supplementary Figure 4 eNpHR3.0-mediated attenuation of evoked EPSC (eEPSC) amplitude is maximal within 50 ms of light onset.
Recordings were performed in acute brain slices from mice expressing eNpHR3.0 in thalamic neurons. (a) Representative voltage clamp recordings of eEPSCs measured in layer 4 of barrel cortex in response to electrical microstimulation of thalamocortical fibers. Depicted are traces for no light (top), light onset 50 ms (middle) and 150 ms (bottom) prior to electrical microstimulation. Time of electrical microstimulation is indicated by black ticks above the traces. (b) Quantification of the efficiency of presynaptic terminal inhibition with increasing delay between light onset and electrical microstimulation. eEPSC amplitude was significantly reduced compared to no-light baseline trials. eEPSC amplitude reduction was greatest during the shortest measured delay of 50 ms (64.3 % reduction compared to no-light baseline; n = 10), significantly more effective than with longer delays. Error bars indicate s.e.m. (* P < 0.05; see Supplementary Table 1 for statistics).
Supplementary Figure 5 Stabilization of intracellular pH strongly attenuates the adaptation of eArch3.0 photocurrents.
(a-b) Example voltage clamp traces of ion pump-expressing neurons before, during and after 5 min constant light application measured in the presence of synaptic blockers (25 µM Ap5, 10 µM CNQX). (a) Buffering of intracellular pH by acetic acid (AcOH) or L-lactate in the external solution stabilized eArch3.0 photocurrents (middle and bottom) compared with sodium-based extracellular solution (top). (b) Example trace of an eNpHR3.0 expressing neuron. Photocurrent adaptation was less pronounced than for eArch3.0. (c) Peak photocurrent amplitudes plotted against steady state photocurrent amplitudes recorded in neurons expressing eNpHR3.0 (n = 25), eArch3.0 (n = 37), or neurons expressing eArch3.0 recorded in the presence of 25 mM AcOH (n = 21) or L-lactate (n = 7). Measurements were fitted with a power law function (f(x) = a*xb+c). The goodness of fit by R-square was 0.73, 0.89, 0.92 and 0.74 for eArch3.0, eArch3.0 in 25 mM AcOH, eArch3.0 in 50 mM L-lactate and eNpHR3.0, respectively. The fits show that with increased light power or expression level, sustained photocurrents greater than 300 pA are possible in the case of the chloride pump eNpHR3.0. However, for the proton pump eArch3.0, photocurrents above ~ 150 pA cannot be maintained over extended periods of time in physiological medium, independent of light power or expression level. This is likely due to changes in internal proton concentration caused by sustained eArch3.0 activation, as buffering of the internal pH leads to attenuation of the photocurrent adaptation. (see Supplementary Table 1 for statistics).
Supplementary Figure 6 Activation of eArch3.0 in cultured hippocampal neurons leads to an increase in intracellular pH that is attenuated by a lactate-containing buffer.
(a) Representative two-photon images depicting mean BCECF fluorescence before (top) and during (bottom) 590 nm illumination in control neurons (left) and in neurons expressing eNpHR3.0 (middle) and eArch3.0 (right). (b) Time course of changes in fluorescence of the intracellular pH indicator BCECF. ΔF/F for non-transduced neurons (black, n = 3 imaged regions), eNpHR3.0-expressing neurons (blue, n = 4 imaged regions), eArch3.0-expressing neurons (red, n = 3 imaged regions) and eArch3.0-expressing neurons recorded in an extracellular buffer containing L-lactate (50 mM, orange, n = 3 imaged regions) under the same light delivery protocol as in Fig. 3. (c) Average ΔF/F during baseline and light-on periods shown in b. (d) Proposed model of eArch3.0-mediated changes in synaptic transmission. Activation of eArch3.0 using yellow light leads to increased intra-synaptic pH, which induces calcium influx into the presynaptic terminal, thus causing increased spontaneous release onto post-synaptic excitatory (red) and inhibitory (blue) neurons. Increased activity of local-circuit interneurons might lead to inhibition onto neurons that do not receive input from the axonal inputs targeted with eArch3.0. Error bars and shaded regions indicate s.e.m. (* P < 0.05; see Supplementary Table 1 for statistics).
Supplementary Figure 7 A chloride-conducting channelrhodopsin inhibits somatic spiking but triggers neurotransmitter release in presynaptic terminals and fails to attenuate evoked release.
(a) Light-evoked outward photocurrents recorded in a cultured neuron transduced with GtACR1 and clamped to 0 mV. Gray value indicates irradiance. Inset: average peak and steady-state photocurrents recorded at a light power density of 2 mW mm–2 (n = 7). (b) Attenuation of spiking in cultured neurons during illumination with increasing light power densities. Neurons were recorded in whole-cell current clamp mode and current injections were performed through the patch pipette to evoke action potentials. The minimal current required to elicit spikes (rheobase) was measured under light pulses of varying light power density. (c) Light-evoked release in GtACR1-expressing thalamocortical terminals. Experiments were conducted in the same configuration as shown in Fig. 2. Depicted are light-evoked EPSC (left) and microstimulation-evoked EPSCs (right). (d) Comparison of light- and microstimulation-evoked EPSCs recorded in the same cells under the same configuration described in c (n = 8). (e) To evaluate the effect of GtACR1 activation in presynaptic terminals, we compared the responses to microstimulation of the thalamocortical fibers following a 10 ms light pulse (top) and under continuous light stimulation (bottom). This was necessary because of the light onset-associated EPSCs, which caused adaptation in the subsequent microstimulation-evoked responses. (f) Summary of GtACR1 experiments, performed as depicted in e (n = 4), showing no effect of light administration on microstimulation-evoked release from thalamocortical terminals. Error bars indicate s.e.m. (see Supplementary Table 1 for statistics).
Supplementary Figures 1–7 and Supplementary Table 1 (PDF 1215 kb)
Cultured hippocampal neurons transduced with an AAV vector encoding SyGCaMP6s were imaged at 0.33 Hz. Spontaneous single-bouton and population-level events are unaffected by yellow light delivery (5 min duration, 590 nm, 2 mW mm-2, indicated by yellow bar). The movie is 30 x accelerated. Imaging area is 295 μm x 295 μm. (AVI 6362 kb)
Cultured hippocampal neurons co-transduced with AAV vectors encoding eArch3.0-mCherry and SyGCaMP6s were imaged at 0.33 Hz. Activation of eArch3.0 (5 min duration, 590nm, 2 mW mm-2, indicated by yellow bar) leads to an increase in SyGCaMP6s fluorescence. The movie is 30 x accelerated. Imaging area is 295 μm x 295 μm. (AVI 21628 kb)
Cultured hippocampal neurons co-transduced with AAV vectors encoding eArch3.0-mCherry and a cytoplasmic GCaMP6s were imaged at 0.33 Hz. Activation of eArch3.0 (5 min duration, 590nm, 2 mW mm-2, indicated by yellow bar) leads to an increase in somatic GCaMP6s fluorescence. The movie is 30 x accelerated. Imaging area is 295 μm x 295 μm. (AVI 12284 kb)
Cultured hippocampal neurons co-transduced with AAV vectors encoding eNpHR3.0-mCherry and SyGCaMP6s were imaged at 0.33 Hz. Activation of eNpHR3.0 (5 min duration, 590nm, 2 mW mm-2, indicated by yellow bar) does not trigger an increase SyGCaMP6s fluorescence. The movie is 30 x accelerated. Imaging area is 295 μm x 295 μm. (AVI 6073 kb)
About this article
Cite this article
Mahn, M., Prigge, M., Ron, S. et al. Biophysical constraints of optogenetic inhibition at presynaptic terminals. Nat Neurosci 19, 554–556 (2016) doi:10.1038/nn.4266
Three Rostromedial Tegmental Afferents Drive Triply Dissociable Aspects of Punishment Learning and Aversive Valence Encoding
Frontiers in Neural Circuits (2019)
Prefrontal cortical ChAT-VIP interneurons provide local excitation by cholinergic synaptic transmission and control attention
Nature Communications (2019)
Nature Neuroscience (2019)
Nature Communications (2019)